Re: [PATCH v4] Input: gpio_keys: Add level trigger support for GPIO keys

On Thu, Mar 1, 2018 at 2:37 PM, Arnd Bergmann [off-list ref] wrote:

On Thu, Mar 1, 2018 at 1:09 PM, Andy Shevchenko
[off-list ref] wrote:

quoted

On Thu, Mar 1, 2018 at 9:36 AM, Baolin Wang [off-list ref] wrote:

quoted

On some platforms (such as Spreadtrum platform), the GPIO keys can only
be triggered by level type. So this patch introduces one trigger_type to
indicate if the button's interrupt type is level trigger or edge trigger.

quoted

                        button->irq =
                                irq_of_parse_and_map(to_of_node(child), 0);

AFAIU, 0 means NO_IRQ, thus,

quoted

+                       if (button->irq)
+                               button->trigger_type =
+                                       irq_get_trigger_type(button->irq);

irq_get_trigger_type(NO_IRQ) should return 0.

Therefore, if (button->irq) is redundant.
Did I miss anything?

"irq_of_parse_and_map(to_of_node(child), 0);" is the first interrupt. A
more common way to express the same thing in modern drivers is
'platform_get_irq(pdev, 0)'.

Yes, though I'm talking about _returned_ value.

However, there is one more thing I'm missing: with my suggestion
of emulating edge-triggered interrupts in the irqchip/gpio driver, the
trigger type here should also be 'IRQF_TRIGGER_RISING |
IRQF_TRIGGER_FALLING' in the DT description, so I don't think
we need to change the gpio-keys driver at all, just the gpio driver.

Sounds even better.

-- 
With Best Regards,
Andy Shevchenko