Created attachment 183321 [details] dmesg output Hi ! I found myself in case described by bug https://bugzilla.kernel.org/show_bug.cgi?id=81691 but with a different laptop. My Thinkpad X240 has a working acpi_video backlight control, but due to bios claiming win8 support, the native backlight interface (i915) is loaded. The native interface works as intended, brightness events are sent to userspace but I have nothing set up to handle them (kde, gnome or acpid ...). Forcing acpi_video to load and handle backlight control, using the quirk described in bug 81691 solves my problem (see patch attached). Thanks, Mathieu
Created attachment 183331 [details] acpidump
Created attachment 183341 [details] proposed patch
(In reply to Mathieu OTHACEHE from comment #0) > Created attachment 183321 [details] > dmesg output > > Hi ! > > I found myself in case described by bug > https://bugzilla.kernel.org/show_bug.cgi?id=81691 but with a different > laptop. > > My Thinkpad X240 has a working acpi_video backlight control, but due to bios > claiming win8 support, the native backlight interface (i915) is loaded. > > The native interface works as intended, brightness events are sent to > userspace but I have nothing set up to handle them (kde, gnome or acpid ...). Is it because you do not use kde/gnome?
Yes, with my distrib : Archlinux+openbox, there's no default userspace program to handle brightness events. I have to use xbindkeys like described here http://www.function.fr/advanced-linux-configuration-for-lenovo-thinkpad-x240/ Just tested with gnome and kde, brightness keys are working out of the box.
If you do not want user space program and prefer the acpi interface, you can add acpi_backlight=video to kernel cmdline on v4.2 kernel or video.use_native_backlight=0 on previous kernel. I think the intention of the DMI table is for systems that its native interface should work but actually didn't and only acpi_video interface works. This isn't the case here and people of this laptop using GNOME/KDE shouldn't have this problem, so I prefer not to add it to the DMI table. Are you OK with this?
I'll close it as documented.
Ok, thanks for your answer. The kernel cmdline solution is working fine !
(In reply to Mathieu OTHACEHE from comment #7) > Ok, thanks for your answer. The kernel cmdline solution is working fine ! Good to know this, thanks!