Hi!
> >@@ -166,10 +174,19 @@ static int omnia_leds_probe(struct i2c_client *client,
> > static int omnia_leds_remove(struct i2c_client *client)
> > {
> >+ u8 buf[5];
> >+
> > /* put all LEDs into default (HW triggered) mode */
> > i2c_smbus_write_byte_data(client, CMD_LED_MODE,
> > CMD_LED_MODE_LED(OMNIA_BOARD_LEDS));
> >+ /* set all LEDs color to [255, 255, 255] */
> >+ buf[0] = CMD_LED_COLOR;
> >+ buf[1] = OMNIA_BOARD_LEDS;
> >+ buf[2] = buf[3] = buf[4] = 255;
> >+
> >+ i2c_master_send(client, buf, 5);
> >+
> > return 0;
> > }
>
> I wonder if we're doing right merging this driver in this form.
> We break the rule one-led-class-device-per-one-channel.
> We don't have LED multi color support yet, so this should support
> RGB LEDs in the old manner. Or switch to using LED multi color
> class.
Fair point.
We treat it as a white LED instead of RGB LED at this
point. One-led-per-channel would be problematic, as hardware
"triggers" are common for all three channels.
So I thought we could go from "white" led to multicolor, when it
becomes available, without going through One-led-per-channel...
I agree this is not quite standard.
Pavel
--
(english) http://www.livejournal.com/~pavelmachek
(cesky, pictures) http://atrey.karlin.mff.cuni.cz/~pavel/picture/horses/blog.html
Attachment:
signature.asc
Description: Digital signature
