Sensel touchpad on a Lenovo laptop breaks on transition to tablet mode

[APerson13]

I recognize that this is almost definitely not a Niri issue, but I have no idea where would be a better place to start asking for help on this. Please direct me to a better space for a question like this if there is one.

The short version of my issue is that if I open my hinge to a point where the laptop transitions to tablet mode and back, my touchpad will stop working usefully until a full power cycle. This happens for sure on both Niri and a Wayland KWin session.

I use a Lenovo Thinkpad X1 Titanium Gen 1 laptop. This laptop has a Sensel-based touchpad, though I can't seem to find much more info than just that online. Linux kernel version is 6.17.0-20-generic.

When the touchpad is in its broken state, libinput still detects it, but debug-events shows some weird behaviour like periodic GESTURE_HOLD events from just keeping my finger mostly still on the touchpad:

-event2   DEVICE_ADDED                 Power Button                      seat0 default group1  cap:k
-event11  DEVICE_ADDED                 Video Bus                         seat0 default group2  cap:k
-event1   DEVICE_ADDED                 Lid Switch                        seat0 default group3  cap:S
-event0   DEVICE_ADDED                 Sleep Button                      seat0 default group4  cap:k
-event7   DEVICE_ADDED                 Wacom Pen and multitouch sensor Finger seat0 default group5  cap:t  size 285x190mm ntouches 10 calib
-event8   DEVICE_ADDED                 Wacom Pen and multitouch sensor Pen seat0 default group5  cap:T  size 285x190mm calib
-event5   DEVICE_ADDED                 SNSL0001:00 2C2F:0006 Touchpad    seat0 default group6  cap:pg  size 85x60mm tap (dl off) left scroll-nat scroll-2fg-edge click-buttonareas-clickfinger dwtp-on
-event4   DEVICE_ADDED                 SNSL0001:00 2C2F:0006 Mouse       seat0 default group6  cap:p left scroll-nat scroll-button
-event3   DEVICE_ADDED                 AT Translated Set 2 keyboard      seat0 default group7  cap:k
-event6   DEVICE_ADDED                 TPPS/2 Elan TrackPoint            seat0 default group8  cap:p left scroll-nat scroll-button
-event10  DEVICE_ADDED                 ThinkPad Extra Buttons            seat0 default group9  cap:kS
-event5   GESTURE_HOLD_BEGIN           +0.040s  1
 event5   GESTURE_HOLD_END             +0.102s  1
 event5   GESTURE_HOLD_BEGIN           +0.802s  1
 event5   GESTURE_HOLD_END             +0.870s  1
 event5   GESTURE_HOLD_BEGIN           +1.577s  1
 event5   GESTURE_HOLD_END             +1.638s  1
 event5   GESTURE_HOLD_BEGIN           +2.352s  1
 event5   GESTURE_HOLD_END             +2.414s  1
 event5   GESTURE_HOLD_BEGIN           +3.127s  1
 event5   GESTURE_HOLD_END             +3.190s  1
 event5   GESTURE_HOLD_BEGIN           +3.902s  1
 event5   GESTURE_HOLD_END             +3.966s  1

Notably, there are no POINTER_MOTION events, effectively making the touchpad unusable. Semi-relatedly, the haptic clicks do not trigger in this state.

From this state, restarting the desktop session, reloading the kernel module i2c_hid and its dependencies (including i2c_hid_acpi, which I will get back to later), and unbinding/rebinding the device through sysfs all fail to bring the touchpad back. I need to fully turn the laptop off, wait a few seconds, and turn it on again to get my touchpad back.

To avoid the issue, I have found 2 methods:

• Use an X11 session (i.e., not Niri, and thus not really desirable to me now that I know what is possible)
• Unload the i2c_hid_acpi kernel module before opening to tablet mode, and load it again when out of tablet mode. I have bound Mod+Shift+O and Mod+Shift+Ctrl+O to these actions respectively for the time being.

When this happens, the most common signal in dmesg is the following 2 messages:

[  111.619149] i2c_designware i2c_designware.0: i2c_dw_handle_tx_abort: lost arbitration
[  112.369247] i2c_hid_acpi i2c-SNSL0001:00: failed to set a report to device: -121

There have been 2 semi-related issues in the past few years, both of which have since been fixed as far as I know:

• Race condition leads to the touchpad sometimes not initializing on boot (fixed in kernel 6.8.9)
• Touchpad sometimes stops working after suspend/resume (I've never experienced this, unsure if it was ever fixed.)

Any guidance on trying to find a more sane workaround for this or trying to fix it properly would be much appreciated.

[Sempyos]

Does the same issue occur on sway or cosmic-comp