| Summary: | KDE-specific tablet mode features not being automatically activated on Lenovo ideapad anymore | ||
|---|---|---|---|
| Product: | [Plasma] kwin | Reporter: | Etaash Mathamsetty <etaash.mathamsetty> |
| Component: | wayland-generic | Assignee: | KWin default assignee <kwin-bugs-null> |
| Status: | RESOLVED UPSTREAM | ||
| Severity: | normal | CC: | kde, nate, notmart |
| Priority: | NOR | Keywords: | regression |
| Version First Reported In: | 5.24.90 | ||
| Target Milestone: | --- | ||
| Platform: | Arch Linux | ||
| OS: | Linux | ||
| Latest Commit: | Version Fixed In: | ||
| Sentry Crash Report: | |||
|
Description
Etaash Mathamsetty
2022-05-28 20:47:22 UTC
> It does disable and renable the touchpad which is strange.
That means that tablet mode is in fact being automatically detected properly by Libinput. Perhaps some signal isn't getting passed onto KWin anymore.
FWIW it's working properly with my Lenovo X1 Yoga G4.
I have a feeling that is caused by lenovo's interesting method of sending the F22 and F23 keys for leaving and entering tablet mode. Could you check if you lenovo computer does this as well? any updates on this? If you go to System Settings > Workspace Behavior > Touch Mode, can you make sure it's set to "Automatically enable as needed"? If it's already set to that, do you see that Touch Mode works properly if you manually set it to "Always enabled" for testing purposes? (In reply to Nate Graham from comment #4) > If you go to System Settings > Workspace Behavior > Touch Mode, can you make > sure it's set to "Automatically enable as needed"? > > If it's already set to that, do you see that Touch Mode works properly if > you manually set it to "Always enabled" for testing purposes? always enabled works fine automatically enabled as needed is set (In reply to Etaash Mathamsetty from comment #5) > (In reply to Nate Graham from comment #4) > > If you go to System Settings > Workspace Behavior > Touch Mode, can you make > > sure it's set to "Automatically enable as needed"? > > > > If it's already set to that, do you see that Touch Mode works properly if > > you manually set it to "Always enabled" for testing purposes? > > always enabled works fine > automatically enabled as needed is set am using plasma 5.25.3 now, along with beta frameworks and it's not working still OK, then I think a KWin or Kirigami person just needs to investigate it. I conveniently have the kwin source code on my machine, (I patched it for zoom screensharing, but don't worry the tablet mode was broken even before I patched it), so if you could point me to the general area to look in the code, I could see where the error is Check out: - src/input.cpp - src/tabletmodemanager.cpp If you're handy with source code and compiling things, you could also to a `git bisect` to find out which exact commit regressed it for you. (In reply to Nate Graham from comment #9) > Check out: > - src/input.cpp > - src/tabletmodemanager.cpp > > If you're handy with source code and compiling things, you could also to a > `git bisect` to find out which exact commit regressed it for you. won't I have to test it for basically every commit to kwin if I go with the git bisect kwin debug console shows the buttons touchpad on and touchpad off being pressed and released but nothing else my suspicions that I stated earlier are correct, my laptop doesn't have a tablet mode switch, instead it has ideapad extra buttons, where it sends a keycode for enabling the touchpad and disabling it (In reply to Etaash Mathamsetty from comment #12) > my suspicions that I stated earlier are correct, my laptop doesn't have a > tablet mode switch, instead it has ideapad extra buttons, where it sends a > keycode for enabling the touchpad and disabling it the code only appears to be checking for the tablet mode switch Were any other packages updated when you upgraded to Plasma 5.25? I ask because it's also possible this is a Libinput regression, that it used may that event to "tablet switch changed" but that broke. this kernel module does the trick!!!!!! https://github.com/lukas-w/yoga-usage-mode Aha, so the issue was in the kernel. Ideally that code would be upstreamed so you don't need a 3rd-party kernel module for this. |