Re: [Mactel-linux-devel] [PATCH] FN key support for aluminum wired Apple keyboards

[M.S. <ma...@su...>]

On Tue, 2007-10-23 at 12:23 +0200, Michel Dänzer wrote:
> On Mon, 2007-10-22 at 23:09 +0200, M.S. wrote:
> > I attached a patch (vs. 2.6.22.9) which enables the hid stuff to
> > recognize the new aluminum wired Apple keyboard's FN key features.
> > http://www.apple.com/keyboard/
> 
> Great keyboard, isn't it? :)
> 

Surprisingly, yes (of course after switching left alt and cmd keys). ;)

> The HID tree at
> git://git.kernel.org/pub/scm/linux/kernel/git/jikos/hid.git already
> carries patches of mine for this. I CC'd this list when submitting those
> but I think they didn't make it here due to some mail setup issues.
> 

Awesome, could not find anything related to the keyboard so I just went
ahead buzzing my own patch. However, I still can't find your patch in
the tree. Would you mind sending me a link so I can compare them?

This list seems to filter your mails in some way.

> Maybe you want to look at those patches and submit any necessary
> followup changes to the HID/input maintainers.
> > Codewise it seems to get a bit messy with all that
> > GEYSER/Powerbook/QUIRKS stuff but I don't know how to make it cleaner at
> > the moment. Comments welcome.
> 
> I re-use the existing quirks and use the device ID to choose the
> translation table.
> 

I did it that way first until I saw that only quirks can be used in
hid-input.c and the product mapping happens in hid-quirks.c (rather than
hardcoding in former).

> 
> > +static struct hidinput_key_translation apple_keyboard[] = {
> > +        { KEY_VOLUMEUP,        KEY_EJECTCD },
> > +        { }
> > +};
> 
> What's this for? Doesn't seem necessary here.
> 

My product is 05ac:0221. The keyboards appears to emit the volumeup key
according USB HID when pressing the eject key. Despite the higher stack
will still work (as you will map whatever with it), I thought this might
need to be corrected so any higher "I map this nice HID key to some
string defaults" get correct keys upfront.

> 
> > -               if (test_bit(usage->code, hid->pb_pressed_numlock) ||
> > -                               test_bit(LED_NUML, input->led)) {
> > +               if (!(hid->quirks & HID_QUIRK_APPLE_KEYBOARD) &&
> > (test_bit(usage->code, hid->pb_pressed_numlock) ||
> > +                               test_bit(LED_NUML, input->led))) {
> >                         trans =
> > find_translation(powerbook_numlock_keys, usage->code);
> 
> I don't understand what this is for either, can you elaborate?
> 
> 

I had issues with the keyboard getting stuck (only FN+FX keys worked) if
this was not "disabled" for this product, however I might just have had
messed something else up while testing.

--- M.S.