Linux 6.1 To Try Enabling HID++ For All Logitech Bluetooth Devices

[phoronix]

Phoronix: Linux 6.1 To Try Enabling HID++ For All Logitech Bluetooth Devices

A change queued in HID-next ahead of the Linux 6.1 cycle aims to enable HID++ protocol usage for all Logitech Bluetooth devices...

https://www.phoronix.com/news/Linux-6.1-Try-Logitech-HID-BT

[Vorpal]

Is there a way to force enable HID++ today on Logitech devices to see if it makes a difference? The kernel module is already loaded for me, but claims it is used by zero devices. Also when I connect this mouse over bluetooth (instead of their unifying reciever) it stops working. That is odd, it worked in the past. Sigh...

EDIT: Figured out the mouse not working at all issue. I was still on a custom kernel that I built to debug a different thing. Didn't have the required module compiled in. Oops.

Still, my original question stands about force enabling HID++

[theriddick]

Well this is probably a good thing since I use a K800 keyboard and a G604 mouse, both USB HID Bluetooth devices.

[blacknova]

The strange problem I have with G915 TKL keyboard is that it keep awakening Linux system from sleep. MX Keys I had before G915 had no such problem.

[theriddick]

The strange problem I have with G915 TKL keyboard is that it keep awakening Linux system from sleep. MX Keys I had before G915 had no such problem.

My K800 has a off/on switch. Maybe yours does too, just get into habit of turning it off when you suspend your PC?

[ChrisIrwin]

Is there a way to force enable HID++ today on Logitech devices to see if it makes a difference? The kernel module is already loaded for me, but claims it is used by zero devices.

Ditto here, reported 0 devices by lsmod. I'm not an input/kmod expert, but I decided to dig into it to see what I found. It appears my device (MX Ergo) is already using hid++, despite the lsmod's reporting.

This is seemingly confirmed with dmesg:

[  708.590481] logitech-hidpp-device 0005:046D:B01D.000F: input,hidraw4: BLUETOOTH HID v0.22 Mouse [Logitech MX Ergo Multi-Device Trackball ] on 08:d2:3e:ec:c6:1e

Long method:

1. Find your device's event number
   

   libinput list-devices

2. Find the event in /sys/devices
   

   $ find /sys/devices -name event27
   	/sys/devices/virtual/misc/uhid/0005:046D:B01D.0010/input/input52/event27​

3. Go to that directory, and check follow the link to the driver:
   

   $ cd /sys/devices/virtual/misc/uhid/0005:046D:B01D.0010/input/input52/event27
   	$ readlink -f device/device/driver
   	/sys/bus/hid/drivers/logitech-hidpp-device​

Why? Probably because this:

$ cat device/device/modalias
hid:b0005g0001v0000046Dp0000B01D​

Is listed in the aliases for hid_logitech_hidpp:

$ modinfo hid_logitech_hidpp | grep b0005
alias: hid:b0005g*v0000046Dp0000B023
alias: hid:b0005g*v0000046Dp0000B01E
v- This one -v
alias: hid:b0005g*v0000046Dp0000B01D
^- This one -^
alias: hid:b0005g*v0000046Dp0000B012
alias: hid:b0005g*v0000046Dp0000B008
alias: hid:b0005g*v0000046Dp0000B30B
alias: hid:b0005g*v0000046Dp0000B309
alias: hid:b0005g*v0000046Dp0000B305
alias: hid:b0005g*v0000046Dp0000B00C​

I'd imagine if your device isn't in that alias list (note, there are a number of wildcard matches), you could maybe do it manually with modprobe?

$ cat /etc/modprobe.d/force-logitech-hidpp.conf
alias hid:b0005g0001v0000046Dp0000B01D hid_logitech_hidpp​

All of the above was checked on Fedora 37 w/ 6.1.5-200, so YMMV.