Created attachment 134282 [details] evemu-record output (up, down, left, right scroll) I initially reported this issue on the mailing list: https://lists.freedesktop.org/archives/xorg/2017-August/058847.html The hardware is "ThinkPad Bluetooth Laser Mouse", and has "4-way scroll" implemented via a normal scroll wheel which tilts left and right onto microswitches (similar to many other devices, including the ubiquitous Logitech B110). I'll summarise it here: -- #0 - The default button mapping has the scroll directions mixed up, sanity is only restored with the following line specified in an xorg.conf.d .conf file: Option "ButtonMapping" "1 2 3 4 5 0 0 6 7" -- #1 - The sideways "clicks" of the wheel don't provide "repeating" events like on similar mice, but instead act like normal mouse buttons (giving a single Press event and a single Release event). This severely cripples the horizontal scroll feature. I was recommended in response to my mailing list post (by Peter Hutterer) to file a bug here along with some extra info, as a Quirk may be necessary. I've attached (or will attach) output evemu-record (with the .conf Option above currently active). I scrolled once up, once down, then clicked once left, held, then released - and did the same once more for right. If there's any more info which would be useful, please don't hesitate to ask.
For clarification, I disabled the Option line I had set, as I wanted to double-check what the default mapping was. The default mapping has the left and right scroll-wheel tilts act as "Forward" and "Back", and there are no more physical buttons remaining to act as the actual scroll. This explains why I effectively shift the mapping 2 buttons down. I thought it best to do that before asked to.
Urgh, ok. we need to add a repeating timer to this one too, just a hwdb quirk won't be enough, this requires some code changes in libinput
I see. Well, in that case, I'll hang tight and just be patient. I can't imagine there are that many others using this same device.
cc-ing benjamin in case the kernel already has infrastructure for this in place (I doubt it, but...). Please attach a dmesg though. Benjamin: hwheel sends button side/extra and does not repeat, so tilting the wheel isnt' very useful - or at least doesn't do what it is supposed to do.
Can I get hid-recorder outputs of the wheel tilt events too? (install hid-replay from http://bentiss.github.io/hid-replay-docs/)
ping?
Sorry, I've been wrapped up in other stuff, and let this slide. I'll try to get this info ASAP.
Closing after over a month in needinfo
Use of freedesktop.org services, including Bugzilla, is subject to our Code of Conduct. How we collect and use information is described in our Privacy Policy.