Re: [PATCH 3/5] input.h: add keycodes for Fn Lock

[Mauro Carvalho Chehab <mchehab@infradead.org>]

Em Mon, 29 Jul 2013 07:03:46 -0300
Mauro Carvalho Chehab <m.chehab@samsung.com> escreveu:

> Hi Dmitry,
> 
> Em Sun, 28 Jul 2013 21:53:58 -0700
> Dmitry Torokhov <dmitry.torokhov@gmail.com> escreveu:
> 
> > Hi Mauro,
> > 
> > On Mon, Jul 29, 2013 at 12:59:37AM -0300, Mauro Carvalho Chehab wrote:
> > > Samsung notebooks have a FN LOCK key. It works like CAPS LOCK or NUM
> > > LOCK keys.
> > > 
> > > When FN LOCK key is pressed, any further press to a key with a blue label
> > > on it (Fn keys) will produce the alternate code.
> > > 
> > > Another press makes the keyboard to return to its normal state.
> > > 
> > > On the notebooks where such feature were found, a FN LOCK on event
> > > produces scancode 0xa8, and a FN LOCK off event produces scancode 0xa9.
> > > 
> > > Yet, it is better to reserve some space at the keymap to allow some
> > > different implementation of this feature where the same keycode might
> > > be used.
> > > 
> > > Also, as this is actually a switch, add a switch indicator to report
> > > when this switch is set/reset.
> > > 
> > > Signed-off-by: Mauro Carvalho Chehab <m.chehab@samsung.com>
> > > ---
> > >  include/uapi/linux/input.h | 5 +++++
> > >  1 file changed, 5 insertions(+)
> > > 
> > > diff --git a/include/uapi/linux/input.h b/include/uapi/linux/input.h
> > > index d584047..4622c34 100644
> > > --- a/include/uapi/linux/input.h
> > > +++ b/include/uapi/linux/input.h
> > > @@ -716,6 +716,10 @@ struct input_keymap_entry {
> > >  #define BTN_DPAD_LEFT		0x222
> > >  #define BTN_DPAD_RIGHT		0x223
> > >  
> > > +#define KEY_FNLOCK_TOGGLE	0x224	/* Request switch Fn on or off */
> > > +#define KEY_FNLOCK_ON		0x225
> > > +#define KEY_FNLOCK_OFF		0x226
> > > +
> > >  #define BTN_TRIGGER_HAPPY		0x2c0
> > >  #define BTN_TRIGGER_HAPPY1		0x2c0
> > >  #define BTN_TRIGGER_HAPPY2		0x2c1
> > > @@ -853,6 +857,7 @@ struct input_keymap_entry {
> > >  #define SW_FRONT_PROXIMITY	0x0b  /* set = front proximity sensor active */
> > >  #define SW_ROTATE_LOCK		0x0c  /* set = rotate locked/disabled */
> > >  #define SW_LINEIN_INSERT	0x0d  /* set = inserted */
> > > +#define SW_FNLOCK		0x0e  /* set = Fn locked */
> > 
> > I am not sure if we need both the keys and the switch, so I would
> > probably simply go with the keys, and not bother with switch. Then we do
> > not need to touch the atkbd driver at all and rely on udev to set up the
> > keymap and force release keys. 
> 
> The hole idea of doing those patches is to have an userspace tool that will
> be able to show software LEDs, like mate-applet-lockeys, that will query
> the input driver to know the current status. So, the better is to keep 
> control of it as soon as kernel starts controlling the Keyboard, as, 
> otherwise, the software LED indicators may be wrong.
> 
> If you think that having both keycodes and a switch is an overkill, then
> the better is to just keep the switch.

After thinking for a little bit, I can see a few alternatives:

1) add patch 3/5 and patch 4/5 as is. No need to apply patch 5/5. The
Fn Lock switch support will start work as soon as udev initializes the
keytables.

2) Add patches 3/5 and patch 4/5 as is, and patch 5/5 rebased, as:
	http://git.infradead.org/users/mchehab/samsung.git/commitdiff/96171bbd42d8ab9dfc3f36e0d4796080a8989344

This way, all keyboards with FN LOCK will require a quirk at atkbd.
The key handling will only start after udev starts.

3) Modify patch 3/5 to only have there a switch (or a LED). That will
require a change at patch 5/5 to use the scancodes to handle the
switch directly.

IMHO, (3) is better, as it will start handling the FN LOCK earlier,
reducing the risk of someone press the key after keyboard reset and
before udev to load the keytable, and providing a more reliable
software LED indication.

Please let me know what works best for you, and I'll redo the patches.

Thanks!
Mauro