You are not logged in.
Hello,
i used to suspend to RAM my laptop, and it worked very well until a few days.
But since the gnome 3.6 update, it randomly (or maybe not so randomly, but i can't reproduce the exacts conditions...) resume to a black screen, no way to log in. i have to ctrl+alt+f2 to get a TTY and restart gdm.
I use gnome 3.6, systemd, and a very common configuration i think.
Does anybody have an idea what can produce this ?
Last edited by effraie (2012-11-17 14:55:41)
Offline
I am also experiencing this issue, and I agree with the OP that it is hard to reproduce.
I also need some ideas as to what could possibly be causing this.
Offline
I did some more systematic test, and it seem that my laptop resume correctly when it suspend to RAM by closing the lid, and fail to resume (black screen with the mouse cursor) when it suspend by being inactive for 10 minutes.
Offline
and here is a log at the moment a failing resume happen : http://pastebin.fr/25770
Offline
Count me in on this one!
Suspend and wake up via user menu works flawlessly, as well as performing 'sudo pm-suspend' and 'systemctl suspend'. But as soon as my system goes to sleep after inactivity, resume fails with exactly the same symptoms as described above.
I'm running Gnome 3.6 with full systemd on a Core i5 2500k Sandy Bridge desktop pc, using the integrated Intel graphics as well as the onboard Intel 82579V Gigabit Network and onboard Intel 6 Series/C200 Series HD Audio. Even though I'm using a full systemd environment for about two month now, the problem never occurred to me on Gnome 3.4, so it's probably related recent changes in Gnome Power Manager.
Does anyone know which suspend method power management uses and why it obviously differs from the one provided by user menu?
I will give you further information on this issue after I've worked myself through the regarding log files.
Last edited by bunghole (2012-11-03 15:00:05)
Offline
i filled a bug there : https://bugs.archlinux.org/task/32407?p … ened=14190
Offline
Hm, I have no fucking clue what the fuck happened, but for some reason, wake up from suspend works now for me.
This is what I recall doing since the last failing attempt to suspend:
- Updating and removing some packages:
[2012-11-02 19:44] Running 'pacman -Syu'
[2012-11-02 19:44] synchronizing package lists
[2012-11-02 19:44] starting full system upgrade
[2012-11-02 19:44] upgraded grantlee (0.2.0-1 -> 0.3.0-1)
[2012-11-02 19:44] >>> To use libvirt as a non-root user:
[2012-11-02 19:44] >>> Use polkit to grant access.
[2012-11-02 19:44] >>> ...or change the access model in /etc/libvirt/libvirtd.conf.
[2012-11-02 19:44] >>> (see unixperms.patch in PKGBUILD for help)
[2012-11-02 19:44]
[2012-11-02 19:44] >>> To finish the upgrade, restart libvirtd by:
[2012-11-02 19:44] >>> 'systemctl restart libvirtd' as root (systemd) or
[2012-11-02 19:44] >>> '/etc/rc.d/libvirtd restart' as root (initscripts) or
[2012-11-02 19:44] >>> rebooting.
[2012-11-02 19:44] >>> You may also need to run 'rm -rf ~/.libvirt'
[2012-11-02 19:44]
[2012-11-02 19:44] upgraded libvirt (0.10.2-4 -> 1.0.0-1)
[2012-11-02 20:09] Running 'pacman -Syu'
[2012-11-02 20:09] synchronizing package lists
[2012-11-02 20:09] starting full system upgrade
[2012-11-02 20:10] >>> Updating module dependencies. Please wait ...
[2012-11-02 20:11] >>> Generating initial ramdisk, using mkinitcpio. Please wait...
[2012-11-02 20:11] ==> Building image from preset: 'default'
[2012-11-02 20:11] -> -k /boot/vmlinuz-linux -c /etc/mkinitcpio.conf -g /boot/initramfs-linux.img
[2012-11-02 20:11] ==> Starting build: 3.6.5-1-ARCH
[2012-11-02 20:11] -> Running build hook: [base]
[2012-11-02 20:11] -> Running build hook: [udev]
[2012-11-02 20:11] -> Running build hook: [autodetect]
[2012-11-02 20:11] -> Running build hook: [pata]
[2012-11-02 20:11] -> Running build hook: [scsi]
[2012-11-02 20:11] -> Running build hook: [sata]
[2012-11-02 20:11] -> Running build hook: [usb]
[2012-11-02 20:11] -> Running build hook: [lvm2]
[2012-11-02 20:11] -> Running build hook: [filesystems]
[2012-11-02 20:11] -> Running build hook: [keymap]
[2012-11-02 20:11] -> Running build hook: [usbinput]
[2012-11-02 20:11] -> Running build hook: [timestamp]
[2012-11-02 20:11] -> Running build hook: [fsck]
[2012-11-02 20:11] ==> Generating module dependencies
[2012-11-02 20:11] ==> Creating gzip initcpio image: /boot/initramfs-linux.img
[2012-11-02 20:11] ==> Image generation successful
[2012-11-02 20:11] ==> Building image from preset: 'fallback'
[2012-11-02 20:11] -> -k /boot/vmlinuz-linux -c /etc/mkinitcpio.conf -g /boot/initramfs-linux-fallback.img -S autodetect
[2012-11-02 20:11] ==> Starting build: 3.6.5-1-ARCH
[2012-11-02 20:11] -> Running build hook: [base]
[2012-11-02 20:11] -> Running build hook: [udev]
[2012-11-02 20:11] -> Running build hook: [pata]
[2012-11-02 20:11] -> Running build hook: [scsi]
[2012-11-02 20:11] -> Running build hook: [sata]
[2012-11-02 20:11] -> Running build hook: [usb]
[2012-11-02 20:11] -> Running build hook: [lvm2]
[2012-11-02 20:11] -> Running build hook: [filesystems]
[2012-11-02 20:11] -> Running build hook: [keymap]
[2012-11-02 20:11] -> Running build hook: [usbinput]
[2012-11-02 20:11] -> Running build hook: [timestamp]
[2012-11-02 20:11] -> Running build hook: [fsck]
[2012-11-02 20:11] ==> Generating module dependencies
[2012-11-02 20:11] ==> Creating gzip initcpio image: /boot/initramfs-linux-fallback.img
[2012-11-02 20:11] ==> Image generation successful
[2012-11-02 20:11] upgraded linux (3.6.4-1 -> 3.6.5-1)
[2012-11-02 20:11] Running 'pacman -U /tmp/yaourt-tmp-bung/PKGDEST.uCP/ttf-roboto-20121102-1-any.pkg.tar.xz'
[2012-11-02 20:11] Updating font cache... done.
[2012-11-02 20:11] upgraded ttf-roboto (20120312-3 -> 20121102-1)
[2012-11-03 08:49] Running 'pacman -Syu'
[2012-11-03 08:49] synchronizing package lists
[2012-11-03 08:49] starting full system upgrade
[2012-11-03 08:49] upgraded libsecret (0.10-1 -> 0.11-1)
[2012-11-03 15:08] Running 'pacman -Syu'
[2012-11-03 15:08] synchronizing package lists
[2012-11-03 15:08] starting full system upgrade
[2012-11-03 15:09] upgraded alacarte (3.6.1-1 -> 3.6.1-2)
[2012-11-03 15:09] upgraded caribou (0.4.4-1 -> 0.4.4-2)
[2012-11-03 15:09] upgraded gnome-applets (3.4.1-2 -> 3.5.92-1)
[2012-11-03 15:09] upgraded gnome-games-extra-data (3.2.0-1 -> 3.2.0-2)
[2012-11-03 15:09] >>> To use libvirt as a non-root user:
[2012-11-03 15:09] >>> Use polkit to grant access.
[2012-11-03 15:09] >>> ...or change the access model in /etc/libvirt/libvirtd.conf.
[2012-11-03 15:09] >>> (see unixperms.patch in PKGBUILD for help)
[2012-11-03 15:09]
[2012-11-03 15:09] >>> To finish the upgrade, restart libvirtd by:
[2012-11-03 15:09] >>> 'systemctl restart libvirtd' as root (systemd) or
[2012-11-03 15:09] >>> '/etc/rc.d/libvirtd restart' as root (initscripts) or
[2012-11-03 15:09] >>> rebooting.
[2012-11-03 15:09] >>> You may also need to run 'rm -rf ~/.libvirt'
[2012-11-03 15:09]
[2012-11-03 15:09] upgraded libvirt (1.0.0-1 -> 1.0.0-2)
[2012-11-03 15:09] upgraded xdg-user-dirs-gtk (0.8-1 -> 0.9-1)
[2012-11-03 16:37] Running 'pacman -S alacarte'
[2012-11-03 16:37] upgraded alacarte (3.6.1-2 -> 3.6.1-2)
[2012-11-03 16:40] Running 'pacman -Rss gnome-games gnome-games-extra-data'
[2012-11-03 16:41] removed gnome-games-extra-data (3.2.0-2)
[2012-11-03 16:41] removed gnome-games (3.6.1-1)
[2012-11-03 16:41] Running 'pacman -Rss aisleriot'
[2012-11-03 16:41] removed aisleriot (3.2.2-1)
[2012-11-03 16:41] removed guile (1.8.8-2)
[2012-11-03 16:41] Running 'pacman -Rss audacious audacious-plugins'
[2012-11-03 16:41] removed audacious (3.3.2-1)
[2012-11-03 16:41] removed libguess (1.1-2)
[2012-11-03 16:41] removed libmowgli (2.0.0-1)
[2012-11-03 16:41] removed audacious-plugins (3.3.2-1)
[2012-11-03 16:42] Running 'pacman -Rss avidemux-gtk avidemux-cli'
[2012-11-03 16:42] removed avidemux-gtk (2.5.6-3)
[2012-11-03 16:42] removed avidemux-cli (2.5.6-3)
[2012-11-03 16:42] Running 'pacman -Rss gnome-boxes'
[2012-11-03 16:42] removed gnome-boxes (3.6.1-1)
[2012-11-03 16:42] removed spice-gtk3 (0.14-1)
[2012-11-03 16:42] removed libvirt-glib (0.1.3-1)
[2012-11-03 16:42] >>> You may wish to delete the group 'libvirt' if it still exists.
[2012-11-03 16:42] >>> It can be removed by running 'groupdel libvirt' as root.
[2012-11-03 16:42]
[2012-11-03 16:42] removed libvirt (1.0.0-2)
[2012-11-03 16:42] removed netcf (0.2.2-1)
[2012-11-03 16:42] removed netcfg (3.0-1)
[2012-11-03 16:42] removed dhcpcd (5.6.2-1)
[2012-11-03 16:42] removed augeas (0.10.0-2)
[2012-11-03 16:42] removed libosinfo (0.2.1-1)
[2012-11-03 16:42] Running 'pacman -Rss wine winetricks'
[2012-11-03 16:42] removed winetricks (20120912-1)
[2012-11-03 16:42] removed cabextract (1.4-2)
[2012-11-03 16:42] removed wine (1.5.16-2)
[2012-11-03 16:42] removed lib32-glu (9.0.0-1)
[2012-11-03 16:42] removed lib32-gettext (0.18.1.1-2)
[2012-11-03 16:42] removed lib32-acl (2.2.51-2)
[2012-11-03 16:43] Running 'pacman -Rss quodlibet'
[2012-11-03 16:43] Running 'pacman -Rss quodlibet quodlibet-plugins'
[2012-11-03 16:43] removed quodlibet-plugins (2.4.1-2)
[2012-11-03 16:43] removed quodlibet (2.4.1-1)
[2012-11-03 16:44] Running 'pacman -Rss gmp'
[2012-11-03 16:46] Running 'pacman -Rss alacarte'
[2012-11-03 16:46] removed alacarte (3.6.1-2)
[2012-11-03 16:46] Running 'pacman -U /tmp/yaourt-tmp-bung/PKGDEST.hr4/alacarte-3.6.1-2-any.pkg.tar.xz'
[2012-11-03 16:46] installed alacarte (3.6.1-2)
[2012-11-03 17:02] Running 'pacman -Syu'
[2012-11-03 17:02] synchronizing package lists
[2012-11-03 17:02] starting full system upgrade
- Resetting most of the dconf schemata in org.gnome.power-manager and org.gnome.settings-daemon.plugins.power
Maybe it was the kernel update? But I'm pretty shure I've tried suspending after that without any success... Weird shit...
UPDATE: Crap! After restarting the computer, wake up from suspend fails again. I'm going through the log files now to see if I can get any hints on what's causing this problem.
Last edited by bunghole (2012-11-03 19:43:55)
Offline
Count me in on this one too. Same symptoms, different hardware:
CPU: Intel T7500;
GPU: ATI Mobility Radeon HD 3650; - OSS driver from repos;
I can confirm that suspending from gnome menu and the suspend key works, but when suspending due to inactivity it seems that GDM cannot resume.
I will try to investigate what is done differently under these two scenarios.
Offline
Same problem here on an i3 + igp.
Same behaviour, if I suspend to ram manually there is no problem at wake up, but if it suspend itself due to inactivity I have the black screen.
Offline
additional up on this one. It started for me after updating to gnome 3.6
Offline
same issue with coreI3 + nvidia ! nothing really usefull in logs ...
Offline
This problem is still persisting and needs more attention.
From the OP's logs, it appears to be a gdm issue, but I've found no solutions.
Offline
Looks like the recent 3.6.2 update fixed this little annoyance.
Can anyone else confirm that waking from suspend after idle works now?
Offline
hum, something strange in your post. my kernel is 3.6.6-1 (and the update is not so recent)
21:25 mathieu@kubrick ~ % pacman -Qi linux
Nom : linux
Version : 3.6.6-1
[...]
Compilé le : lun. 05 nov. 2012 12:06:33 CET
Installé le : mer. 07 nov. 2012 00:03:13 CET
[...]
Offline
I wasn't talking about the kernel but about many GNOME components which might have caused the issue (like gnome-shell, gnome-settings-daemon, etc.) being updated to 3.6.2...
Update: Three resumes from suspend later, still no system crash. In fact, suspend works better than ever now on my machine, since the few seconds delay I was facing with GNOME 3.4 are gone as well.
Last edited by bunghole (2012-11-15 14:47:00)
Offline
it's ok for me too
Offline
Jap, the issue definitely seems to be fixed now. So unless anyone objects, you might wanna mark this thread (as well as your bug report) as [solved]...
Big thx btw to all the GNOME devs for the rather quick reaction! Just imagine such an issue on generic Windows systems; at first, Microsoft would probably deny that there is problem after all, then after three month release a bugfix which solves wake up, but now prevents the system from going to suspend...
Last edited by bunghole (2012-11-16 20:51:12)
Offline
Yep, confirming that the issue is fixed!
Offline