GNOME Bugzilla – Bug 745753
After recovering from standby keyboard shortcuts don't work
Last modified: 2015-04-14 18:30:02 UTC
I don't shut my PC down often (I just put it in standby mode). Lately I noticed that when I recover from suspending all keyboard shortcuts stop working (CTRL+ALT+T which I set up to open gnome-terminal, and even the default ones which adjust the volume). I have not investigated thoroughly, so I'm not sure "how many recovers" are needed. If necessary I will try to reproduce the bug again.
When this happens please run 'xev -event keyboard' and attach the output you get on the terminal here when pressing one of those shortcuts.
Created attachment 298737 [details] Output of "xev -event keyboard"
Comment on attachment 298737 [details] Output of "xev -event keyboard" This is the output when pressing the following shortcuts: CTRL+ALT+T Volume Up (fn + F12) Volume Down (fn + F11) Volume UP (fn + F12)
Ok, we still have a grab. What's the output of $ dbus-monitor --session --profile | grep ^sig when you press a shortcut in this state?
Created attachment 298739 [details] Output of "dbus-monitor --session --profile | grep ^sig" (the same shortcuts as described before)
Cool, can you try running gnome-settings-daemon from a terminal like: $ /usr/libexec/gnome-settings-daemon --replace --debug > /tmp/gsd.log 2>&1 & The exact path to gnome-settings-daemon might be different on your system. Restarting it, by itself, might make the shortcuts start working so you might have to get again into that state while capturing this log.
Indeed, that command made shortcuts work again. So I suspended 3 or 4 times (none of which broke the shorcuts), then I suspended another time for some time (more than half an hour) and this time only CLTR+ALT+T broke. Anyway, when I do the Volume Up/Down shortcut I see that some logs get appended to /tmp/gsd.log, like: > (gnome-settings-daemon:14058): media-keys-plugin-DEBUG: Launching action for key type '5' (on device id 3) > (gnome-settings-daemon:14058): media-keys-plugin-DEBUG: Could not find device node for XInput device 3 > (gnome-settings-daemon:14058): Gvc-DEBUG: lookup-device-from-stream found device: device description 'Cuffie analogiche', device port = 'analog-output-headphones', device stream id 1 AND stream port = 'analog-output-headphones' stream id '1' and stream description 'Audio interno Stereo analogico' > (gnome-settings-daemon:14058): Gvc-DEBUG: gvc_mixer_control_lookup_device_from_stream - Could not find a device for stream 'Audio interno Stereo analogico' When I do CTRL+ALT+T, instead, nothing happens to the file.
I reproduced the "complete" issue (no shorcut working) and I can confirm that nothing gets logged to the /tmp/gsd.log file when I press them. The only things that work are the keyboard backlight enabler, fn+F6 (which works but doesn't log anything), the screen backlight up/down button (fn+F4/F5), which log something like: > (gnome-settings-daemon:5980): media-keys-plugin-DEBUG: Launching action for key type '56' (on device id 3) > (gnome-settings-daemon:5980): power-plugin-DEBUG: Calling method 'org.gnome.SettingsDaemon.Power.Screen.StepUp' for Power > (gnome-settings-daemon:5980): power-plugin-DEBUG: screen step up > (gnome-settings-daemon:5980): media-keys-plugin-DEBUG: Launching action for key type '56' (on device id 3) > (gnome-settings-daemon:5980): power-plugin-DEBUG: Calling method 'org.gnome.SettingsDaemon.Power.Screen.StepUp' for Power > (gnome-settings-daemon:5980): power-plugin-DEBUG: screen step up > (gnome-settings-daemon:5980): media-keys-plugin-DEBUG: Launching action for key type '56' (on device id 3) > (gnome-settings-daemon:5980): power-plugin-DEBUG: Calling method 'org.gnome.SettingsDaemon.Power.Screen.StepUp' for Power > (gnome-settings-daemon:5980): power-plugin-DEBUG: screen step up > (gnome-settings-daemon:5980): power-plugin-DEBUG: Received gnome session inhibitor change > (gnome-settings-daemon:5980): power-plugin-DEBUG: setting up sleep callback 3600s > (gnome-settings-daemon:5980): media-keys-plugin-DEBUG: Launching action for key type '57' (on device id 3) > (gnome-settings-daemon:5980): power-plugin-DEBUG: Calling method 'org.gnome.SettingsDaemon.Power.Screen.StepDown' for Power > (gnome-settings-daemon:5980): power-plugin-DEBUG: screen step down > (gnome-settings-daemon:5980): media-keys-plugin-DEBUG: Launching action for key type '57' (on device id 3) > (gnome-settings-daemon:5980): power-plugin-DEBUG: Calling method 'org.gnome.SettingsDaemon.Power.Screen.StepDown' for Power > (gnome-settings-daemon:5980): power-plugin-DEBUG: screen step down > (gnome-settings-daemon:5980): media-keys-plugin-DEBUG: Launching action for key type '57' (on device id 3) > (gnome-settings-daemon:5980): power-plugin-DEBUG: Calling method 'org.gnome.SettingsDaemon.Power.Screen.StepDown' for Power > (gnome-settings-daemon:5980): power-plugin-DEBUG: screen step down the WiFi disabler/enabler (fn+F2) which logs this: > (gnome-settings-daemon:5980): rfkill-plugin-DEBUG: RFKILL event: idx 0 type 1 (WLAN) op 2 (CHANGE) soft 0 hard 1 > (gnome-settings-daemon:5980): rfkill-plugin-DEBUG: RFKILL event: idx 12 type 2 (RFKILL) op 1 (DEL) soft 1 hard 0 > (gnome-settings-daemon:5980): sharing-plugin-DEBUG: current network: > (gnome-settings-daemon:5980): sharing-plugin-DEBUG: current network name: > (gnome-settings-daemon:5980): sharing-plugin-DEBUG: conn type: > (gnome-settings-daemon:5980): sharing-plugin-DEBUG: status: 0 > (gnome-settings-daemon:5980): rfkill-plugin-DEBUG: RFKILL event: idx 0 type 1 (WLAN) op 2 (CHANGE) soft 1 hard 1 and the monitor setting button (fn+F1) which logs a list of monitor configuration (the logs mentions that fn+F7 is being handled, though it's actually fn+F1 on my laptop, not sure if that's relevant)
I'm testing GNOME 3.16 and now I can't reproduce the bug anymore. Some days have passed and I haven't seen this bug anymore... maybe it's fixed. I'll update if the bug happens again.
So, this is working correctly? Which distro are you using?
Archlinux. Yes, the issue has disappeared since I updated to GNOME 3.16.
OK, let's close this then - please reopen if the issue turns up again.
I'm experiencing the same bug. I opened a separate report with bug 747745.