Re: [PATCH 1/1] pinctrl: stmfx: add irq_request/release_resources callbacks

[Amelie DELAUNAY]

Hi Linus,

On 10/5/19 6:49 PM, Linus Walleij wrote:
> On Fri, Oct 4, 2019 at 2:29 PM Amelie Delaunay <amelie.delaunay@xxxxxx> 
> wrote:
>
> > When an STMFX IO is used as interrupt through the interrupt-controller
> > binding, the STMFX driver should configure this IO as input. Default
> > value of STMFX IO direction is input, but if the IO is used as output
> > before the interrupt use, it will not work without these callbacks.
> >
> > Signed-off-by: Amelie Delaunay <amelie.delaunay@xxxxxx>
>
> OK I see what you want to achieve.
>
> > +static int stmfx_gpio_irq_request_resources(struct irq_data *data)
> > +{
> > +ÂÂÂÂÂÂ struct gpio_chip *gpio_chip = irq_data_get_irq_chip_data(data);
> > +ÂÂÂÂÂÂ struct stmfx_pinctrl *pctl = gpiochip_get_data(gpio_chip);
> > +ÂÂÂÂÂÂ int ret;
> > +
> > +ÂÂÂÂÂÂ ret = stmfx_gpio_direction_input(&pctl->gpio_chip, data->hwirq);
> > +ÂÂÂÂÂÂ if (ret)
> > +ÂÂÂÂÂÂÂÂÂÂÂÂÂÂ return ret;
> > +
> > +ÂÂÂÂÂÂ ret = gpiochip_lock_as_irq(&pctl->gpio_chip, data->hwirq);
> > +ÂÂÂÂÂÂ if (ret) {
> > +ÂÂÂÂÂÂÂÂÂÂÂÂÂÂ dev_err(pctl->dev, "Unable to lock gpio %lu as IRQ: %d\n",
> > +ÂÂÂÂÂÂÂÂÂÂÂÂÂÂÂÂÂÂÂÂÂÂ data->hwirq, ret);
> > +ÂÂÂÂÂÂÂÂÂÂÂÂÂÂ return ret;
> > +ÂÂÂÂÂÂ }
> > +
> > +ÂÂÂÂÂÂ return 0;
> > +}
>
> Just call gpiochip_reqres_irq() instead of calling the lock etc
> explicitly.
>
> > +static void stmfx_gpio_irq_release_resources(struct irq_data *data)
> > +{
> > +ÂÂÂÂÂÂ struct gpio_chip *gpio_chip = irq_data_get_irq_chip_data(data);
> > +ÂÂÂÂÂÂ struct stmfx_pinctrl *pctl = gpiochip_get_data(gpio_chip);
> > +
> > +ÂÂÂÂÂÂ gpiochip_unlock_as_irq(&pctl->gpio_chip, data->hwirq);
> > +}
>
> Just call gpiochip_relres_irq()
>
> But all this duplicated a lot of code from the core which is not so nice.
>
> > @@ -678,6 +706,8 @@ static int stmfx_pinctrl_probe(struct platform_device *pdev)
> > ÂÂÂÂÂÂÂÂ pctl->irq_chip.irq_set_type = stmfx_pinctrl_irq_set_type;
> > ÂÂÂÂÂÂÂÂ pctl->irq_chip.irq_bus_lock = stmfx_pinctrl_irq_bus_lock;
> > ÂÂÂÂÂÂÂÂ pctl->irq_chip.irq_bus_sync_unlock = stmfx_pinctrl_irq_bus_sync_unlock;
> > +ÂÂÂÂÂÂ pctl->irq_chip.irq_request_resources = stmfx_gpio_irq_request_resources;
> > +ÂÂÂÂÂÂ pctl->irq_chip.irq_release_resources = stmfx_gpio_irq_release_resources;
>
> What about just adding
>
> pctl->irq_chip.irq_enable and do stmfx_gpio_direction_input()
> in that callback instead? gpiolib will helpfully wrap it.

Thanks for pointing that out to me.

I can't use .irq_enable because of I2C transfer to set gpio direction 
(scheduling while atomic BUG occurs in this case). Indeed, .irq_enable 
is called under raw_spin_lock_irqsave in __setup_irq() while 
irq_request_resources is called before.

I could apply gpio direction in stmfx_pinctrl_irq_bus_sync_unlock 
depending on pctl->irq_gpi_src[] (checking which one is set, to set 
input direction), but this will be applied each time a consumer requests 
a stmfx gpio irq.

IMHO, keeping .irq_request/release_resources callbacks and using 
gpiochip_reqres_irq()/gpiochip_relres_irq() seems to be the best compromise.

Regards,
Amelie