On Tue, 1 Feb 2022 at 13:19, Marc Zyngier <[email protected]> wrote:
>
> Move the reference to the device over to the irq domain.
>
> Signed-off-by: Marc Zyngier <[email protected]>
> ---
> drivers/pinctrl/pinctrl-starfive.c | 3 ++-
> 1 file changed, 2 insertions(+), 1 deletion(-)
>
> diff --git a/drivers/pinctrl/pinctrl-starfive.c b/drivers/pinctrl/pinctrl-starfive.c
> index 0b912152a405..5be9866c2b3c 100644
> --- a/drivers/pinctrl/pinctrl-starfive.c
> +++ b/drivers/pinctrl/pinctrl-starfive.c
> @@ -1307,7 +1307,6 @@ static int starfive_probe(struct platform_device *pdev)
> sfp->gc.base = -1;
> sfp->gc.ngpio = NR_GPIOS;
>
> - starfive_irq_chip.parent_device = dev;
> starfive_irq_chip.name = sfp->gc.label;
>
> sfp->gc.irq.chip = &starfive_irq_chip;
> @@ -1330,6 +1329,8 @@ static int starfive_probe(struct platform_device *pdev)
> if (ret)
> return dev_err_probe(dev, ret, "could not register gpiochip\n");
>
> + irq_domain_set_pm_device(sfp->gc.irq.domain, dev);
> +
The gpio framework uses the irq_domain at sfp->gc.irq.domain, so
shouldn't this be set before registering the gpio_chip with
devm_gpiochip_add_data above?
> out_pinctrl_enable:
> return pinctrl_enable(sfp->pctl);
> }
> --
> 2.30.2
>
On Tue, 01 Feb 2022 15:16:39 +0000,
Emil Renner Berthing <[email protected]> wrote:
>
> On Tue, 1 Feb 2022 at 13:19, Marc Zyngier <[email protected]> wrote:
> >
> > Move the reference to the device over to the irq domain.
> >
> > Signed-off-by: Marc Zyngier <[email protected]>
> > ---
> > drivers/pinctrl/pinctrl-starfive.c | 3 ++-
> > 1 file changed, 2 insertions(+), 1 deletion(-)
> >
> > diff --git a/drivers/pinctrl/pinctrl-starfive.c b/drivers/pinctrl/pinctrl-starfive.c
> > index 0b912152a405..5be9866c2b3c 100644
> > --- a/drivers/pinctrl/pinctrl-starfive.c
> > +++ b/drivers/pinctrl/pinctrl-starfive.c
> > @@ -1307,7 +1307,6 @@ static int starfive_probe(struct platform_device *pdev)
> > sfp->gc.base = -1;
> > sfp->gc.ngpio = NR_GPIOS;
> >
> > - starfive_irq_chip.parent_device = dev;
> > starfive_irq_chip.name = sfp->gc.label;
> >
> > sfp->gc.irq.chip = &starfive_irq_chip;
> > @@ -1330,6 +1329,8 @@ static int starfive_probe(struct platform_device *pdev)
> > if (ret)
> > return dev_err_probe(dev, ret, "could not register gpiochip\n");
> >
> > + irq_domain_set_pm_device(sfp->gc.irq.domain, dev);
> > +
>
> The gpio framework uses the irq_domain at sfp->gc.irq.domain, so
> shouldn't this be set before registering the gpio_chip with
> devm_gpiochip_add_data above?
It is devm_gpiochip_add_data() that create the domain, so there is
nothing to set before.
Thanks,
M.
--
Without deviation from the norm, progress is not possible.
On Tue, 1 Feb 2022 at 16:16, Emil Renner Berthing <[email protected]> wrote:
> On Tue, 1 Feb 2022 at 13:19, Marc Zyngier <[email protected]> wrote:
> >
> > Move the reference to the device over to the irq domain.
> >
> > Signed-off-by: Marc Zyngier <[email protected]>
> > ---
> > drivers/pinctrl/pinctrl-starfive.c | 3 ++-
> > 1 file changed, 2 insertions(+), 1 deletion(-)
> >
> > diff --git a/drivers/pinctrl/pinctrl-starfive.c b/drivers/pinctrl/pinctrl-starfive.c
> > index 0b912152a405..5be9866c2b3c 100644
> > --- a/drivers/pinctrl/pinctrl-starfive.c
> > +++ b/drivers/pinctrl/pinctrl-starfive.c
> > @@ -1307,7 +1307,6 @@ static int starfive_probe(struct platform_device *pdev)
> > sfp->gc.base = -1;
> > sfp->gc.ngpio = NR_GPIOS;
> >
> > - starfive_irq_chip.parent_device = dev;
> > starfive_irq_chip.name = sfp->gc.label;
> >
> > sfp->gc.irq.chip = &starfive_irq_chip;
> > @@ -1330,6 +1329,8 @@ static int starfive_probe(struct platform_device *pdev)
> > if (ret)
> > return dev_err_probe(dev, ret, "could not register gpiochip\n");
> >
> > + irq_domain_set_pm_device(sfp->gc.irq.domain, dev);
> > +
>
> The gpio framework uses the irq_domain at sfp->gc.irq.domain, so
> shouldn't this be set before registering the gpio_chip with
> devm_gpiochip_add_data above?
Ah, no. sfp->gc.irq.domain is a pointer to an irq_domain that is
initialised when adding the gpio_chip.
Reviewed-by: Emil Renner Berthing <[email protected]>