Hello Doug,

On 08/08/2014 10:54 PM, Doug Anderson wrote:

Hi,

quoted

To fix the issue a variation of patch [0] will be posted that moves the IRQ
pinmux setup from .irq_set_type to the .irq_request_resources function handler.
That way the pin will be setup as IRQ regardless of the the trigger type [1]
when someone calls request_[threaded]_irq().

Only the mentioned patch fixes the issue but Tomasz said that even a call to
gpio_direction_{input,output} can change the pin configuration so he will post
another patch that will add a bit mask to samsung_pin_bank to prevent any pinmux
reconfiguration.

Would just making a device tree change fix this?  AKA for the pin, do:

  samsung,pin-function = <0xf>;

Yes, when configuring the pin as 0xf (GPIO interrupt) instead of 0x0 (GPIO
input) the issue is not present indeed. So after Nick answer my question about
what I got wrong with the "linux,gpio-keymap" property, I will change the pin
function when re-posing the DTS changes for the atmel touchpad.

I have some vague recollection that I set interrupts to pin-function
"0" by default for some reason (assuming they would go to 0xf when
interrupts were enabled).  ...but I can't for the life of me remember
if it was a good reason or just seemed like the right thing to do.

It would be great to know if there is a good reason. I see indeed that all
pinctrl setup in the Peach Pit/Pi and Snow DTS that involves interrupts are
using 0x0 as the pin function. Since as Tomasz explained Samsung SoC makes a
difference between GPIO-IRQ and GPIO input I guess that it's better to change
those to 0xf instead. What do you think?

Regardless of this though I think that both the patch to move the IRQ
pinmux setup from .irq_set_type to the .irq_request_resources and the patch to
to prevent any pinmux reconfiguration are good improvements to avoid future
issues like the one we found here.

-Doug

Best regards,
Javier