Re: [PATCH 1/4] add gpio keyboard support

[Jean-Christophe PLAGNIOL-VILLARD <plagnioj@jcrosoft.com>]

> 
> This information would be far more informative if you printed the
> button that is pressed/released here.
ok
> 
> > +		}
> > +		gb->previous_state = val;
> > +	}
> 
> You do nothing for debouncing the keys.
yes my hardware manage directly (gpio IP)

we can add it later
> > +			return ret;
> > +		}
> > +		gpio_direction_input(gpio);
> > +	}
> > +
> > +	pdata->poller.func = gpio_key_poller;
> > +
> > +	cdev = &pdata->cdev;
> > +	dev->type_data = cdev;
> > +	cdev->dev = dev;
> > +	cdev->f_caps = CONSOLE_STDIN;
> > +	cdev->tstc = gpio_keys_tstc;
> > +	cdev->getc = gpio_keys_getc;
> > +
> > +	console_register(&pdata->cdev);
> 
> As I tried to explain yesterday I'm not sure that directly registering
> the gpio keys as a console is the right thing to do. A keyboard has
> more attributes than what we have in the console. For example a keyboard
> has a button press / button release event which we don't have on the
> console.
agreed but today we do not have a input framework yet

I'm planning to add such framework but as a second step
I'll add the usb keyboard too
> 
> > +
> > +	return poller_register(&pdata->poller);
> > +}
> > +
> > +static struct driver_d gpio_keys_driver = {
> > +	.name	= "gpio_keys",
> > +	.probe	= gpio_keys_probe,
> > +};
> > +
> > +static int gpio_keys_init(void)
> > +{
> > +	register_driver(&gpio_keys_driver);
> > +	return 0;
> 
> Since we now do not panic anymore when an initcall fails but only
> use the return value for debug purposes please do a
> 
> return register_driver(&gpio_keys_driver);
ok

Best Regards,
J.

_______________________________________________
barebox mailing list
barebox@lists.infradead.org
http://lists.infradead.org/mailman/listinfo/barebox