Re: [PATCH v3 3/3] pinctrl: cherryview: Pass irqchip when adding gpiochip

[Hans de Goede <hdegoede@redhat.com>]

Hi,

On 13-11-2019 20:27, Andy Shevchenko wrote:
> On Wed, Nov 13, 2019 at 08:05:20PM +0100, Hans de Goede wrote:
>> We need to convert all old gpio irqchips to pass the irqchip
>> setup along when adding the gpio_chip. For more info see
>> drivers/gpio/TODO.
>>
>> For chained irqchips this is a pretty straight-forward conversion.
>>
>> Acked-by: Mika Westerberg <mika.westerberg@linux.intel.com>
>> Signed-off-by: Hans de Goede <hdegoede@redhat.com>
> 
>>   	struct irq_chip irqchip;
>>   	void __iomem *regs;
>> +	unsigned int irq;
>>   	unsigned intr_lines[16];
> 
> This will conflict with our for-next.

Ah, I did cherry-pick intel-pinctrl for-next into my tree a couple of days
back, but I see there is a new "pinctrl: cherryview: Missed type change to unsigned int"
commit there which causes this conflict. I have cherry picked this new
commit into my tree and I will send out a v4 which should not conflict.

> 
>> +	if (need_valid_mask)
>> +		chip->irq.init_valid_mask = chv_init_irq_valid_mask;
>> +	chip->irq.init_hw = chv_gpio_irq_init_hw;
>> +	chip->irq.parent_handler = chv_gpio_irq_handler;
>> +	chip->irq.num_parents = 1;
>> +	chip->irq.parents = &pctrl->irq;
>> +	chip->irq.default_type = IRQ_TYPE_NONE;
>> +	chip->irq.handler = handle_bad_irq;
>>   
>>   	if (!need_valid_mask) {
>>   		irq_base = devm_irq_alloc_descs(pctrl->dev, -1, 0,
> 
> 
> Perhaps now it makes sense to
> 
> 	if (need_valid_mask) {
> 		chip->irq.init_valid_mask = chv_init_irq_valid_mask;
> 	} else {
> 		irq_base = devm_irq_alloc_descs(pctrl->dev, -1, 0,
> 		...
> 	} >
> ?

Ack good one, will also change this for v4.

Regards,

Hans

p.s.

About upstreaming this, I know this has a pre-requisite on the new add_ranges
callback stuff, but how about Linus Walleij creating an immutable branch
of his tree with the first series which adds the add_ranges callback in
there and then you merge that branch into pinctrl-intel/for-next and then
we just upstream all of this for 5.5 ? That seems easier then spreading
it out over 2 cycles. Just my 2 cents.