Bug 423175 - krunner not invoked by global shortcut (changed from default)
Summary: krunner not invoked by global shortcut (changed from default)
Status: RESOLVED WORKSFORME
Alias: None
Product: systemsettings
Classification: Applications
Component: kcm_keys (other bugs)
Version First Reported In: 5.18.5
Platform: Fedora RPMs Linux
: NOR normal
Target Milestone: ---
Assignee: Michael Jansen
URL:
Keywords:
Depends on:
Blocks:
 
Reported: 2020-06-18 17:27 UTC by Mark T. Kennedy
Modified: 2024-09-03 03:47 UTC (History)
6 users (show)

See Also:
Latest Commit:
Version Fixed/Implemented In:
Sentry Crash Report:


Attachments

Note You need to log in before you can comment on or make changes to this bug.
Description Mark T. Kennedy 2020-06-18 17:27:04 UTC
SUMMARY

After changing the global shortcut designed to invoke krunner from Alt-Space/Alt-F2 to Meta-F2, I can't invoke krunner by pressing Meta-F2.  With the input focus set to the desktop root window, I can still invoke krunner by typing printable characters like "xterm" or "emacs".  


STEPS TO REPRODUCE

1. Use the System Settings app to modify the global shortcut defaults for invoking krunner from Alt-Space and Alt-F2 to just Meta-F2.
2. Type Meta-F2

OBSERVED RESULT

Krunner does not drop down.  It is not started automatically if it isn't already running.


EXPECTED RESULT

The krunner one line input window should drop down from the top of the screen.


SOFTWARE/OS VERSIONS

Linux/KDE Plasma: Fedora 32 Workstation Edition
KDE Plasma Version: 5.18.5
KDE Frameworks Version: 5.70.0
Qt Version: 5.14.2

ADDITIONAL INFORMATION

This behavior was persistently observed after using xmodmap to change my physical Alt_L key to emit the Meta_L keysym event.  The physical Alt_R key was unchanged and still emits Alt_R.
Comment 1 Oussema Bouaneni 2020-06-22 14:45:22 UTC
I am not able to reproduce this on my computer. Krunner starts fine for me using Meta-F2.
Here's my system information if it helps:
Operating System: Manjaro Linux 
KDE Plasma Version: 5.18.5
KDE Frameworks Version: 5.70.0
Qt Version: 5.15.0
Kernel Version: 5.7.0-3-MANJARO
Comment 2 David Edmundson 2020-06-22 15:26:32 UTC
>This behavior was persistently observed after using xmodmap to change my physical Alt_L key to emit the Meta_L keysym event.

So you can no longer reproduce with that step?
Comment 3 Mark T. Kennedy 2020-06-22 15:48:50 UTC
no, it persists across reboots.  reliably.  xev verifies that Meta-F2 is being generated.  and krunner ignores it.  if you want to explore this with me further, my time is available.  and i'm old and very experienced developer.
Comment 4 Christoph Feck 2020-06-29 20:17:43 UTC
Requested information was added with comment 2; changing status for inspection.
Comment 5 Mark T. Kennedy 2020-07-05 20:12:55 UTC
chris - what does your last comment mean?  this behavior is persistent.
Comment 6 Mark T. Kennedy 2020-07-12 23:32:43 UTC
to recap - the Alt_L physical key on my keyboard generates the Meta_L X Windows keysym.  the Alt_R physical key generates the Alt_R keysym.  System Settings for global shortcuts shows that invoking krunner is bound to Meta-F2 with no other binding.  But typing Meta-F2 does *nothing*.  If i add a shortcut binding for a bare F2 or Shift-F2 to invoke krunner, they *work*.  strangely, i noticed tonight that the left and right 'Windows' keys, which i have configured to generate the X Windows keysyms Super_L and Super_R respectively, invoke krunner!  yet as i said earlier, the only binding shown in System Settings is for Meta-F2.  this situation is *persistent* and deterministic and repeatable.  if it helps, this is the output of 'xmodmap -pm':

% xmodmap -pm
xmodmap:  up to 2 keys per modifier, (keycodes in parentheses):

shift       Shift_L (0x32),  Shift_R (0x3e)
lock      
control     Control_L (0x25),  Control_R (0x69)
mod1        Meta_L (0x40),  Meta_L (0xcd)
mod2        Alt_R (0x6c)
mod3        Super_L (0x85),  Super_L (0xce)
mod4        Hyper_L (0x42),  Hyper_L (0xcf)
mod5        ISO_Level3_Shift (0x5c),  Mode_switch (0xcb)
Comment 7 Mark T. Kennedy 2020-07-15 20:33:15 UTC
ah, so if kwin *is* the process that owns the events that trigger krunner, then i discovered something relevant today.  i was trying to invoke a gnu emacs action via Meta-<mouse drag> and instead the emacs window was selected and i was able to move it.  surprised, i went to System Settings and discovered that this behavior was supposed to be on *Alt*, not *Meta*.  recall that i used xmodmap to remap Alt to be Meta.  so apparently kwin *ignores* xmodmap changes and doesn't use the current X server keysym.

thoughts/comments?
Comment 8 Mark T. Kennedy 2020-07-18 15:05:09 UTC
(In reply to Oussema Bouaneni from comment #1)
> I am not able to reproduce this on my computer. Krunner starts fine for me
> using Meta-F2.
> Here's my system information if it helps:
> Operating System: Manjaro Linux 
> KDE Plasma Version: 5.18.5
> KDE Frameworks Version: 5.70.0
> Qt Version: 5.15.0
> Kernel Version: 5.7.0-3-MANJARO

are you using a Meta key that was assigned *before* KDE was started?  or did you use xmodmap to change the modifier setup before trying to use it?
Comment 9 Mark T. Kennedy 2020-07-18 22:22:51 UTC
i've been doing some reading and have a question. what matters to KDE/Qt?  the state bit returned in a keypress event (essentially the alias of a slot in the modifier table) or the keysym associated with the keycode used to trigger the turning on of the state bit? emacs, for example, cares about the *keysym*.  the System Settings app (/usr/bin/systemsettings5) cares about the *keysym*. but i'm guessing kwin actually cares about the modifier map slot and has a hard-coded understanding that mod1 is Alt and mod2 is Meta?
Comment 10 Alexander Lohnau 2020-07-19 06:38:54 UTC
Seems related to BUG 424266
Comment 11 Mark T. Kennedy 2020-07-19 17:04:26 UTC
yes, it has to be the same.  i have assigned Alt_L to mod1 again and changed the global shortcut to invoke krunner back to Alt-F2.  it does *not* work directly after a reboot but, as in the other bug report, if i restart krunner it, it *does* work.  so strange.
Comment 12 Alexander Lohnau 2020-07-19 18:08:30 UTC
>as in the other bug report, if i restart krunner it, it *does* work.
Do you mean krunner or kglobalaccel5? In the mentioned bug report kglobalaccel5 was restarted.
Comment 13 Mark T. Kennedy 2020-07-19 18:28:27 UTC
kglobalaccel5!  sorry, not enough morning coffee :-).
Comment 14 fanzhuyifan 2024-08-03 01:23:11 UTC
Is this still an issue as of plasma 6.1? A lot has changed since the plasma 5 times and maybe the issue has been fixed.
Comment 15 Bug Janitor Service 2024-08-19 03:47:24 UTC
๐Ÿ›๐Ÿงน โš ๏ธ This bug has been in NEEDSINFO status with no change for at least 15 days. Please provide the requested information, then set the bug status to REPORTED. If there is no change for at least 30 days, it will be automatically closed as RESOLVED WORKSFORME.

For more information about our bug triaging procedures, please read https://community.kde.org/Guidelines_and_HOWTOs/Bug_triaging.

Thank you for helping us make KDE software even better for everyone!
Comment 16 Bug Janitor Service 2024-09-03 03:47:27 UTC
๐Ÿ›๐Ÿงน This bug has been in NEEDSINFO status with no change for at least 30 days. Closing as RESOLVED WORKSFORME.