[1/1] irqchip: stm32-exti: add domain translate function

Message ID 1547215132-85612-1-git-send-email-loic.pallardy@st.com
State New
Headers show
Series
  • [1/1] irqchip: stm32-exti: add domain translate function
Related show

Commit Message

Loic PALLARDY Jan. 11, 2019, 1:58 p.m.
Domain translate function is needed to recover irq
configuration parameters from DT node

Signed-off-by: Loic Pallardy <loic.pallardy@st.com>

---
 drivers/irqchip/irq-stm32-exti.c | 18 ++++++++++++++++++
 1 file changed, 18 insertions(+)

-- 
2.7.4

Comments

Marc Zyngier Jan. 11, 2019, 3:52 p.m. | #1
On 11/01/2019 13:58, Loic Pallardy wrote:
> Domain translate function is needed to recover irq

> configuration parameters from DT node

> 

> Signed-off-by: Loic Pallardy <loic.pallardy@st.com>


Surely this fixes something, right? Can you please add a Fixes: tag?

> ---

>  drivers/irqchip/irq-stm32-exti.c | 18 ++++++++++++++++++

>  1 file changed, 18 insertions(+)

> 

> diff --git a/drivers/irqchip/irq-stm32-exti.c b/drivers/irqchip/irq-stm32-exti.c

> index 6edfd4bfa169..717879028cc4 100644

> --- a/drivers/irqchip/irq-stm32-exti.c

> +++ b/drivers/irqchip/irq-stm32-exti.c

> @@ -819,9 +819,27 @@ static int __init stm32_exti_init(const struct stm32_exti_drv_data *drv_data,

>  	return ret;

>  }

>  

> +static int stm32_exti_h_translate(struct irq_domain *d,

> +				  struct irq_fwspec *fwspec,

> +				  unsigned long *out_hwirq,

> +				  unsigned int *out_type)

> +{

> +	if (is_of_node(fwspec->fwnode)) {


Is there any case where this wouldn't be true? i.e. an stm32 ACPI system?

> +		if (fwspec->param_count != 2)

> +			return -EINVAL;

> +

> +		*out_hwirq = fwspec->param[0];

> +		*out_type = fwspec->param[1] & IRQ_TYPE_SENSE_MASK;

> +		return 0;

> +	}

> +

> +	return -EINVAL;

> +}


Because otherwise, this is exactly what irq_domain_xlate_twocell() does.

> +

>  static const struct irq_domain_ops stm32_exti_h_domain_ops = {

>  	.alloc	= stm32_exti_h_domain_alloc,

>  	.free	= irq_domain_free_irqs_common,

> +	.translate = stm32_exti_h_translate,

>  };

>  

>  static int

> 


Thanks,

	M.
-- 
Jazz is not dead. It just smells funny...
Loic PALLARDY Jan. 11, 2019, 5:34 p.m. | #2
Hi Marc,
Thanks for the review

> -----Original Message-----

> From: Marc Zyngier <marc.zyngier@arm.com>

> Sent: vendredi 11 janvier 2019 16:52

> To: Loic PALLARDY <loic.pallardy@st.com>; tglx@linutronix.de;

> jason@lakedaemon.net; mcoquelin.stm32@gmail.com; Alexandre TORGUE

> <alexandre.torgue@st.com>

> Cc: linux-kernel@vger.kernel.org; linux-stm32@st-md-

> mailman.stormreply.com; linux-arm-kernel@lists.infradead.org

> Subject: Re: [PATCH 1/1] irqchip: stm32-exti: add domain translate function

> 

> On 11/01/2019 13:58, Loic Pallardy wrote:

> > Domain translate function is needed to recover irq

> > configuration parameters from DT node

> >

> > Signed-off-by: Loic Pallardy <loic.pallardy@st.com>

> 

> Surely this fixes something, right? Can you please add a Fixes: tag?

Yes I will add tag of patch introducing hierarchical domain

> 

> > ---

> >  drivers/irqchip/irq-stm32-exti.c | 18 ++++++++++++++++++

> >  1 file changed, 18 insertions(+)

> >

> > diff --git a/drivers/irqchip/irq-stm32-exti.c b/drivers/irqchip/irq-stm32-

> exti.c

> > index 6edfd4bfa169..717879028cc4 100644

> > --- a/drivers/irqchip/irq-stm32-exti.c

> > +++ b/drivers/irqchip/irq-stm32-exti.c

> > @@ -819,9 +819,27 @@ static int __init stm32_exti_init(const struct

> stm32_exti_drv_data *drv_data,

> >  	return ret;

> >  }

> >

> > +static int stm32_exti_h_translate(struct irq_domain *d,

> > +				  struct irq_fwspec *fwspec,

> > +				  unsigned long *out_hwirq,

> > +				  unsigned int *out_type)

> > +{

> > +	if (is_of_node(fwspec->fwnode)) {

> 

> Is there any case where this wouldn't be true? i.e. an stm32 ACPI system?


You're right, no ACPI so check useless

> 

> > +		if (fwspec->param_count != 2)

> > +			return -EINVAL;

> > +

> > +		*out_hwirq = fwspec->param[0];

> > +		*out_type = fwspec->param[1] & IRQ_TYPE_SENSE_MASK;

> > +		return 0;

> > +	}

> > +

> > +	return -EINVAL;

> > +}

> 

> Because otherwise, this is exactly what irq_domain_xlate_twocell() does.


Thanks for pointing me this helper. Indeed no need to duplicate code
I'll send a v2 

Regards,
Loic
> 

> > +

> >  static const struct irq_domain_ops stm32_exti_h_domain_ops = {

> >  	.alloc	= stm32_exti_h_domain_alloc,

> >  	.free	= irq_domain_free_irqs_common,

> > +	.translate = stm32_exti_h_translate,

> >  };

> >

> >  static int

> >

> 

> Thanks,

> 

> 	M.

> --

> Jazz is not dead. It just smells funny...

Patch

diff --git a/drivers/irqchip/irq-stm32-exti.c b/drivers/irqchip/irq-stm32-exti.c
index 6edfd4bfa169..717879028cc4 100644
--- a/drivers/irqchip/irq-stm32-exti.c
+++ b/drivers/irqchip/irq-stm32-exti.c
@@ -819,9 +819,27 @@  static int __init stm32_exti_init(const struct stm32_exti_drv_data *drv_data,
 	return ret;
 }
 
+static int stm32_exti_h_translate(struct irq_domain *d,
+				  struct irq_fwspec *fwspec,
+				  unsigned long *out_hwirq,
+				  unsigned int *out_type)
+{
+	if (is_of_node(fwspec->fwnode)) {
+		if (fwspec->param_count != 2)
+			return -EINVAL;
+
+		*out_hwirq = fwspec->param[0];
+		*out_type = fwspec->param[1] & IRQ_TYPE_SENSE_MASK;
+		return 0;
+	}
+
+	return -EINVAL;
+}
+
 static const struct irq_domain_ops stm32_exti_h_domain_ops = {
 	.alloc	= stm32_exti_h_domain_alloc,
 	.free	= irq_domain_free_irqs_common,
+	.translate = stm32_exti_h_translate,
 };
 
 static int