<html> <head> <base href="https://bugs.freedesktop.org/" /> </head> <body> <div> <a class="bz_bug_link bz_status_UNCONFIRMED " title="UNCONFIRMED - libinput keep open motion sensor input device" href="https://bugs.freedesktop.org/show_bug.cgi?id=91563#c6">Comment # 6</a> on <a class="bz_bug_link bz_status_UNCONFIRMED " title="UNCONFIRMED - libinput keep open motion sensor input device" href="https://bugs.freedesktop.org/show_bug.cgi?id=91563">bug 91563</a> from <a class="email" href="mailto:dev@mgautier.fr" title="Matthieu Gautier <dev@mgautier.fr>"> Matthieu Gautier</a> <pre>Created <a href="attachment.cgi?id=118007" name="attach_118007" title="patch to add to linux tree, to add INPUT_PROP_ACCELEROMETER">attachment 118007</a> <a href="attachment.cgi?id=118007&action=edit" title="patch to add to linux tree, to add INPUT_PROP_ACCELEROMETER">[details]</a> patch to add to linux tree, to add INPUT_PROP_ACCELEROMETER Sorry for my silence but, as said Bastien, I was away. (And, as I'm leaving again, I'll be silent for the next few weeks..) Indeed, the tag INPUT_PROP_ACCELEROMETER is missing in applesmc. Adding it [1] doesn't fix the problem. It shows that it is more complex that it seems ... With the tag INPUT_PROP_ACCELEROMETER, a recent systemd version (>219, the fedora22 version[2]) properly add the udev ID_INPUT_ACCELEROMETER=1 tag. With this tag, libinput still open the device. However, it seems to be the right thing to do. Gnome(-shell?) now detect the device is a accelerometer and display a button to deactivate screen rotation, so we should keep the device open. So I've add a custom udev rules to remove the udev ID_INPUT (and ID_INPUT_ACCELEROMETER) tag: SUBSYSTEM=="input" ATTRS{properties}=="40" ENV{ID_INPUT}="" ENV{ID_INPUT_ACCELEROMETER}="" Without a ID_INPUT, libinput should ignore the device, but.. it doesn't. The device is still open, but it seems to be never closed. Each time, I go back to gdm (close user session), the device is open one more time. lsof report each time that the gnome-shell process open N time the same device. I have still not found why. Once I'll go back, I'll investigate this again. [1] I haven't submit my patch to kernel, but just apply the attached patch to linux tree if you what to test. [2] If you what to test with 219 systemd version, you can add a custom rules to add the ID_INPUT_ACCELEROMETER yourself : SUBSYSTEM=="input" ATTRS{properties}=="40" ENV{ID_INPUT_ACCELEROMETER}="1"</pre> </div> <hr> You are receiving this mail because: <ul> <li>You are the assignee for the bug.</li> </ul> </body> </html>