Bug 480091

Summary:	kdeconnectd dies upon connection
Product:	[Applications] kdeconnect	Reporter:	Jona Wagner <jona>
Component:	common	Assignee:	Albert Vaca Cintora <albertvaka>
Status:	RESOLVED FIXED
Severity:	normal	CC:	andrew.g.r.holmes, nicolas.fella, smowtenshi
Priority:	NOR
Version First Reported In:	23.08.4
Target Milestone:	---
Platform:	Arch Linux
OS:	Linux
Latest Commit:		Version Fixed/Implemented In:
Sentry Crash Report:
Attachments:	GDB logs

Description Jona Wagner 2024-01-20 11:21:38 UTC

Created attachment 165071 [details]
GDB logs

SUMMARY
kdeconnectd process stops upon receiving a connection from my android phone, other devices (tested with an iPad) seem to be working.


STEPS TO REPRODUCE
This isn't 100% reproducible, more like 90%, it happens pretty much daily on boot up. I can fix it by restarting kdeconnectd often enough, eventually it just starts working.

OBSERVED RESULT
kdeconnectd keeps stopping with SIGTERM

EXPECTED RESULT
kdeconnectd works


SOFTWARE/OS VERSIONS
Linux/KDE Plasma: 
(available in About System)
KDE Plasma Version: 5.27.10
KDE Frameworks Version: 5.114.0
Qt Version: 5.15.12

ADDITIONAL INFORMATION

I did some debugging. First let me describe my debug method:
For debugging: `gdb /usr/lib/kdeconnectd`
>(gdb) catch signal all
>Catchpoint 1 (any signal)
>(gdb) commands 1
>Type commands for breakpoint(s) 1, one per line.
>End with a line saying just "end".
>>bt
>>continue
>>end

When the process exits gracefully (because KDEconnectd is already started by a systemd unit named `dbus-:1.3-org.kde.kdeconnect@<counter>.service`): `pkill kdeconnectd` and restart kdeconnectd in gdb.

I did a traces with both my android phone which stops kdeconnectd and my iPad which works correctly.
When connecting the android phone kdeconnectd seems to perform a fork, which kills it. Please see the attachment for the trace(s).

Comment 1 Nicolas Fella 2024-01-26 13:01:46 UTC

Can you the logs (journalctl _COMM=kdeconnectd) from right before the problem? It looks like a plugin fails to load because of library issues

Comment 2 Jona Wagner 2024-01-26 13:35:52 UTC

(In reply to Nicolas Fella from comment #1)
> Can you the logs (journalctl _COMM=kdeconnectd) from right before the
> problem? It looks like a plugin fails to load because of library issues

These are the journalctl logs from today. kdeconnectd is no longer running

» journalctl -b _COMM=kdeconnectd
Jan 26 09:15:45 hostname kdeconnectd[1982]: new capabilities for  "Android"
Jan 26 09:15:45 hostname kdeconnectd[1982]: kdeconnect.plugin.battery: No Primary Battery detected on this system. This may be a bug.
Jan 26 09:15:45 hostname kdeconnectd[1982]: kdeconnect.plugin.battery: Total quantity of batteries found:  0
Jan 26 09:15:45 hostname kdeconnectd[1982]: Connecting to deprecated signal QDBusConnectionInterface::serviceOwnerChanged(QString,QString,QString)
Jan 26 09:15:45 ryzen-mini kdeconnectd[1982]: kf.modemmanagerqt: Failed enumerating MM objects: "org.freedesktop.DBus.Error.NameHasNoOwner" 
                                               "Could not activate remote peer: activation request failed: unknown unit."
Jan 26 09:58:41 hostname kdeconnectd[1982]: qt.qpa.wayland: Creating a fake screen in order for Qt not to crash
Jan 26 13:15:42 hostname kdeconnectd[1982]: new capabilities for  "Android"
Jan 26 13:15:42 hostname kdeconnectd[1982]: kdeconnect.plugin.battery: No Primary Battery detected on this system. This may be a bug.
Jan 26 13:15:42 hostname kdeconnectd[1982]: kdeconnect.plugin.battery: Total quantity of batteries found:  0


see also:

» journalctl -b | grep kdeconnect
Jan 26 09:15:44 hostname systemd[1511]: Created slice Slice /app/dbus-:1.3-org.kde.kdeconnect.
Jan 26 09:15:44 hostname systemd[1511]: Started dbus-:1.3-org.kde.kdeconnect@0.service.
Jan 26 09:15:45 hostname kdeconnectd[1982]: new capabilities for  "Android"
Jan 26 09:15:45 hostname kdeconnectd[1982]: kdeconnect.plugin.battery: No Primary Battery detected on this system. This may be a bug.
Jan 26 09:15:45 hostname kdeconnectd[1982]: kdeconnect.plugin.battery: Total quantity of batteries found:  0
Jan 26 09:15:45 hostname kdeconnectd[1982]: Connecting to deprecated signal QDBusConnectionInterface::serviceOwnerChanged(QString,QString,QString)
Jan 26 09:15:45 hostname kdeconnectd[1982]: kf.modemmanagerqt: Failed enumerating MM objects: "org.freedesktop.DBus.Error.NameHasNoOwner" 
Jan 26 09:58:41 hostname kdeconnectd[1982]: qt.qpa.wayland: Creating a fake screen in order for Qt not to crash
Jan 26 13:03:42 hostname plasmashell[1766]: file:///usr/share/plasma/plasmoids/org.kde.kdeconnect/contents/ui/FullRepresentation.qml:37:23: QML DeviceDelegate: Binding loop detected for property "width"
Jan 26 13:15:42 hostname kdeconnectd[1982]: new capabilities for  "Android"
Jan 26 13:15:42 hostname kdeconnectd[1982]: kdeconnect.plugin.battery: No Primary Battery detected on this system. This may be a bug.
Jan 26 13:15:42 hostname kdeconnectd[1982]: kdeconnect.plugin.battery: Total quantity of batteries found:  0
Jan 26 13:15:49 hostname systemd[1511]: dbus-:1.3-org.kde.kdeconnect@0.service: Consumed 10.769s CPU time.
Jan 26 13:15:49 hostname plasmashell[1766]: file:///usr/share/plasma/plasmoids/org.kde.kdeconnect/contents/ui/FullRepresentation.qml:37:23: QML DeviceDelegate: Binding loop detected for property "width"

Comment 3 Jona Wagner 2024-01-26 13:38:49 UTC

Might also be some kind of dbus issue, see this log:

» journalctl -b | grep "26 13:15"
Jan 26 13:15:42 hostname kdeconnectd[1982]: new capabilities for  "Android"
Jan 26 13:15:42 hostname kdeconnectd[1982]: kdeconnect.plugin.battery: No Primary Battery detected on this system. This may be a bug.
Jan 26 13:15:42 hostname kdeconnectd[1982]: kdeconnect.plugin.battery: Total quantity of batteries found:  0
Jan 26 13:15:49 hostname dbus-broker[1539]: Monitor :1.256 is being disconnected as it attempted to send a message.
Jan 26 13:15:49 hostname systemd[1511]: dbus-:1.3-org.kde.kdeconnect@0.service: Consumed 10.769s CPU time.
Jan 26 13:15:49 hostname plasmashell[1766]: file:///usr/share/plasma/plasmoids/org.kde.kdeconnect/contents/ui/FullRepresentation.qml:37:23: QML DeviceDelegate: Binding loop detected for property "width"

Comment 4 Jona Wagner 2024-03-08 11:08:41 UTC

looks like this issue got fixed with the KDE Plasma 6 release (kdeconnect 24.02.0)