[PATCH V4 Resend 2/3] mfd: stmpe: Remove irq_trigger from platform data

Viresh Kumar viresh.kumar at linaro.org
Thu Nov 29 21:01:49 EST 2012


On 29 November 2012 15:04, Lee Jones <lee.jones at linaro.org> wrote:
> On Thu, 29 Nov 2012, Viresh Kumar wrote:
>
>> STMPE can confige
>
>             configure?
>
>> the way the device emits interrupts and till now this
>
>                                           until?

Ahh... Will fix them. This happens when you send patches at midnight. :)

>> diff --git a/drivers/mfd/stmpe.c b/drivers/mfd/stmpe.c

>>               ret = devm_request_threaded_irq(ci->dev, stmpe->irq, NULL,
>> -                             stmpe_irq, pdata->irq_trigger | IRQF_ONESHOT,
>> -                             "stmpe", stmpe);
>> +                             stmpe_irq, IRQF_ONESHOT, "stmpe", stmpe);
>
> Forgive my ignorance, but you're no longer passing irq_trigger.
>
> Is this intentional? If so, why was it needed before and not now?

Yes, it was intentional. I thought it wasn't required at all. But my mind
is changing a bit now. I feel it is not required for DT, as trigger prop
is already passed in the interrupts cell. But for non-DT user, this is
still required. As there is not other way by which IRQ controller will
come to know what irq trigger type to enable for this irq line.

--
viresh


More information about the devicetree-discuss mailing list