You are not logged in.
Hi All,
Ever since the last full system upgrade (which I performed around 18th of May) I keep getting random disconnects from my Logitech MX Anywhere 2s mouse (connected via its own receiver, not Bluetooth). Every time that happens, I get in the system log the following two lines, then, after a 2 seconds glitch, the mouse reconnects:
May 27 14:11:39 amd org_kde_powerdevil[1269]: qt.core.qobject.connect: QObject::disconnect: Unexpected nullptr parameter
May 27 14:11:39 amd org_kde_powerdevil[1269]: qt.core.qobject.connect: QObject::disconnect: Unexpected nullptr parameterA grep in journalctl's output for "Unexpected nullptr parameter" also yields a couple of appearances of the same error coming from konsole.
My question here is if & how can I further drill down for relevant details so I can open a bug report to the QT guys (I guess).
Thanks in advance.
Last edited by mdcclxv (2024-05-27 18:57:28)
Offline
I don't think it's a QT bug, perhaps the disconnection triggered a bug in KDE or QT library, thus showing the error message, but it's not the reason that caused the disconnection.
Perhaps you can try
killall org_kde_powerdevilto rule out the possibility that it's powerdevil that disconnected your mouse.
Other than these, I have no idea...or try record the time when your mouse disconnects, and look up the error messages in `sudo journalctl -b 0` and `sudo dmesg -H`?
Offline
"Qt" - and no, it's not a Qt bug for sure.
QObject::disconnect is part of Qt's signal and slot system (callback functions) and there's apparently a bug in powerdevil but that's not disconnecting your input device.
You're looking for some powersaving thing, maybe https://wiki.archlinux.org/title/Power_ … utosuspend ("usbcore.autosuspend=-1" will disable it but nb. that userspace tools like TLP can and will override that at runtime and you've to configure it there!)
Other than that: does the mouse disconnect mid-movement or is only slow to respond after you ignored it for a while?
In that case, does clicking any of the buttons make it re-connect faster?
Offline
I don't think it's a QT bug, perhaps the disconnection triggered a bug in KDE or QT library, thus showing the error message, but it's not the reason that caused the disconnection.
Perhaps you can trykillall org_kde_powerdevilto rule out the possibility that it's powerdevil that disconnected your mouse.
Other than these, I have no idea...or try record the time when your mouse disconnects, and look up the error messages in `sudo journalctl -b 0` and `sudo dmesg -H`?
Not sure why boot stuff has any relevance, because my issue is happening during system usage, long after the boot is finished.
Offline
"Qt" - and no, it's not a Qt bug for sure.
QObject::disconnect is part of Qt's signal and slot system (callback functions) and there's apparently a bug in powerdevil but that's not disconnecting your input device.You're looking for some powersaving thing, maybe https://wiki.archlinux.org/title/Power_ … utosuspend ("usbcore.autosuspend=-1" will disable it but nb. that userspace tools like TLP can and will override that at runtime and you've to configure it there!)
I've added the following rule in /etc/udev/rules.d/, but I'm getting an "Invalid key/value pair, ignoring" error. Is there any reference of valid udev keys for each subsystem?
ACTION=="add", SUBSYSTEM=="usb", TEST=="power/control", ATTR{power/control}="auto", ATTR{power/autosuspend}=-1Other than that: does the mouse disconnect mid-movement or is only slow to respond after you ignored it for a while?
In that case, does clicking any of the buttons make it re-connect faster?
Now that you've asked me this, I find that the mouse stays disconnected for as long as I don't touch it and reconnects in a split second when I move it.
Last edited by mdcclxv (2024-05-27 14:56:09)
Offline
That's most likely the mouse's powersaving.
The rule there is wrong.
ATTR{power/control}="auto"does the straight opposite of what you'd want and "usbcore.autosuspend=-1" is a https://wiki.archlinux.org/title/Kernel_parameters
Offline
That's most likely the mouse's powersaving.
The rule there is wrong.ATTR{power/control}="auto"does the straight opposite of what you'd want and "usbcore.autosuspend=-1" is a https://wiki.archlinux.org/title/Kernel_parameters
I got that usbcore.autosuspend is a kernel parameter, but I don't have a boot-loader so I would have to mess with EFI boot entries. I figured I would first try a udev rule, since the "power/autosuspend" attribute is documented in udev wiki.
I tried the following rule (using Logitech's receiver usb vendor&device IDs), still the same error of "Invalid key/value pair". Looks to me like the "Power Management" wiki is outdated on the "power/autosuspend" attribute part.
ACTION=="add", SUBSYSTEM=="usb", ATTR{idVendor}=="046d", ATTR{idProduct}=="c52b", ATTR{power/autosuspend}=-1I will add the usbcore.autosuspend kernel param and come back with the result.
Offline
ATTR{power/control}="on"Idk whether the "-1" works here as "invalid" delay to disable it but you'd have to quote it.
Offline
My bad, I totally missed the quoting mistake ![]()
So far it looks like the rule bellow does the trick:
ACTION=="add", SUBSYSTEM=="usb", ATTR{idVendor}=="046d", ATTR{idProduct}=="c52b", ATTR{power/autosuspend}="-1"So the wiki is up-to-date ![]()
Ok, so there's a workaround for me to use the mouse, but the "QObject::disconnect: Unexpected nullptr parameter" issue still remains. I will mark this topic as solved and open a bug on powerdevil project.
Thanks again(!) for your help ![]()
Offline
archlinux org_kde_powerdevil[1672]: qt.core.qobject.connect: QObject::disconnect: Unexpected nullptr parameterI have the same problem and I couldn't use my bluetooth connected keyboard. After disable the autosuspend in the configuration file /etc/tlp.conf by set
# Set to 0 to disable, 1 to enable USB autosuspend feature.
# Default: 1
USB_AUTOSUSPEND=0 I can use my keyboard now, while the errors continue emerge in the log.
Offline