RE: [PATCH 1/3] platform/x86: dell-wmi: add new backlight events

[Mario.Limonciello]

> -----Original Message-----
> From: Pali Rohár <pali@xxxxxxxxxx>
> Sent: Monday, June 8, 2020 3:36 PM
> To: Limonciello, Mario
> Cc: y.linux@xxxxxxxxxxxxx; linux-kernel@xxxxxxxxxxxxxxx; platform-driver-
> x86@xxxxxxxxxxxxxxx; mjg59@xxxxxxxxxxxxx
> Subject: Re: [PATCH 1/3] platform/x86: dell-wmi: add new backlight events
> 
> 
> [EXTERNAL EMAIL]
> 
> On Monday 08 June 2020 20:14:15 Mario.Limonciello@xxxxxxxx wrote:
> > > >>> index c25a4286d766..0b4f72f923cd 100644
> > > >>> --- a/drivers/platform/x86/dell-wmi.c
> > > >>> +++ b/drivers/platform/x86/dell-wmi.c
> > > >>> @@ -252,6 +252,10 @@ static const struct key_entry
> > > >> dell_wmi_keymap_type_0010[] = {
> > > >>>  	/* Fn-lock switched to multimedia keys */
> > > >>>  	{ KE_IGNORE, 0x1, { KEY_RESERVED } },
> > > >>>
> > > >>> +	/* Backlight brightness level */
> > > >>> +	{ KE_KEY,    0x57, { KEY_BRIGHTNESSDOWN } },
> > > >>> +	{ KE_KEY,    0x58, { KEY_BRIGHTNESSUP } },
> > > >>> +
> > > >
> > > > For these particular events are they emitted by another interface as
> well
> > > in this
> > > > platform?
> > > >
> > > > If so they should be KE_IGNORE so you don't end up with double
> > > notifications to
> > > > userspace.
> > > Thank you both for the review,
> > > This is my first patch so if i am doing something wrong please let me
> know.
> > >
> > > Both before and after this change they are only emitted once (verified
> via
> > > showkeys)
> > > this is because `dell_wmi_process_key()` calls
> > > `acpi_video_handles_brightness_key_presses()`
> > > for brightness events, and
> `acpi_video_handles_brightness_key_presses()`
> > > makes sure no duplicate acpi-video events are sent.
> >
> > That's good to hear that it also filters it, but my opinion is that dell-
> wmi.c
> > should also filter it.  So just change KE_KEY to KE_IGNORE like the other
> events.
> 
> IIRC for other existing KEY_BRIGHTNESS* lines, KE_KEY/KE_IGNORE decision
> is there just because kernel can be configured if ACPI layer would
> handle them or not. And based on
> acpi_video_handles_brightness_key_presses()
> function we know if ACPI layer processed these keys or not.
> 
> So in my opinion we should handle these new KEY_BRIGHTNESS* events in
> the same way. So dell-wmi should process these events, but only in case
> when ACPI layer did not processed them.

OK thanks, I wasn't aware of this decision elsewhere.