Re: [PATCH] Support for Snapstream Firefly remote added to ati_remote.c

[Vojtech Pavlik]

On Tue, Sep 21, 2004 at 11:35:55AM -0700, dylan@xxxxxxxxxxxxxx wrote:

> I've added support for the Snapstream Firefly X10 remote
> (http://www.snapstream.com/Products/firefly/). This involved adding a new
> product ID, and adding some additional logic to the event_lookup function.
> The Firefly alternately encodes the keypress data differently than the
> driver currently expects, so that every other keypress is invalid with the
> current driver version. The transformation is pretty simple; I'm assuming
> it's there simply to allow the driver to distinguish between what two
> distinct key events looks like and what it looks like to drop signal in
> the middle of holding a remote button.
> 
> As well, I also added a module_param called disable_keyboard which allows
> the user to toggle off whether the driver actually sends key events, or
> simply generates raw evdev events. I found this useful, since I'm working
> on an app that reads the evdev events natively and then sends it's own
> keyboard/mouse/window events in X. The keyboard events being sent by the
> driver were interfering with the events I wanted to generate myself.
> 
> This patch applies to version 2.6.9-rc1-mm4.
> 
> Signed-off-by: Dylan Paris <kernelcontrib@xxxxxxxxxxxxxx>
> 
> @@ -280,6 +289,9 @@ static struct
>         {KIND_FILTERED, 0xf4, 0x2F, EV_KEY, KEY_END, 1},        /* END */
>         {KIND_FILTERED, 0xf5, 0x30, EV_KEY, KEY_SELECT, 1},     /* SELECT */
> 
> +       /* Firefly Remote Buttons */
> +       {KIND_FILTERED, 0xf1, 0x2c, EV_KEY, KEY_KATAKANA, 1},     /* MAXIMIZE */
> +

Now, now, would it be too hard to add a proper key definition into
input.h? I don't think you want the remote to change character sets on
Japanese machines.

> @@ -460,6 +479,7 @@ static void ati_remote_input_report(stru
>         struct input_dev *dev = &ati_remote->idev;
>         int index, acc;
>         int remote_num;
> +       int ev_type;
> 
>         /* Deal with strange looking inputs */
>         if ( (urb->actual_length != 4) || (data[0] != 0x14) ||
> @@ -492,7 +512,9 @@ static void ati_remote_input_report(stru
> 
>         if (ati_remote_tbl[index].kind == KIND_LITERAL) {
>                 input_regs(dev, regs);
> -               input_event(dev, ati_remote_tbl[index].type,
> +
> +               ev_type = ((disable_keyboard) ? 0 : > ati_remote_tbl[index].type);
> +               input_event(dev, ev_type,
>                         ati_remote_tbl[index].code,
>                         ati_remote_tbl[index].value);
>                 input_sync(dev);

I won't let you abuse the input API this way. Event type 0 is EV_SYN and
is reserved for synchronization and configuration change notifications.
Definitely not for sending events that look like keystrokes but aren't.

There is a nice ioctl, called EVIOCGRAB which will give you what you
want (the console won't be receiving the events anymore) without any
ugly hacks.

-- 
Vojtech Pavlik
SuSE Labs, SuSE CR
-
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo@xxxxxxxxxxxxxxx
More majordomo info at  http://vger.kernel.org/majordomo-info.html
Please read the FAQ at  http://www.tux.org/lkml/