Re: [PATCH 07/35] ARM: davinci: aintc: use irq domain

From: Bartosz Golaszewski
Date: Tue Feb 05 2019 - 11:29:22 EST


pon., 4 lut 2019 o 23:42 David Lechner <david@xxxxxxxxxxxxxx> napisaÅ(a):
>
> On 1/31/19 7:39 AM, Bartosz Golaszewski wrote:
> > From: Bartosz Golaszewski <bgolaszewski@xxxxxxxxxxxx>
> >
> > We need to create an irq domain if we want to select SPARSE_IRQ. The
> > cp-intc driver already supports it, but aintc doesn't. Use the helpers
> > provided by the generic irq chip abstraction.
> >
> > Signed-off-by: Bartosz Golaszewski <bgolaszewski@xxxxxxxxxxxx>
> > ---
> > arch/arm/mach-davinci/irq.c | 38 ++++++++++++++++++++++++++-----------
> > 1 file changed, 27 insertions(+), 11 deletions(-)
> >
> > diff --git a/arch/arm/mach-davinci/irq.c b/arch/arm/mach-davinci/irq.c
> > index e539bc65d4ef..c874ea269411 100644
> > --- a/arch/arm/mach-davinci/irq.c
> > +++ b/arch/arm/mach-davinci/irq.c
> > @@ -23,6 +23,7 @@
> > #include <linux/interrupt.h>
> > #include <linux/irq.h>
> > #include <linux/io.h>
> > +#include <linux/irqdomain.h>
> >
> > #include <mach/hardware.h>
> > #include <mach/cputype.h>
> > @@ -43,6 +44,7 @@
> > #define IRQ_INTPRI7_REG_OFFSET 0x004C
> >
> > static void __iomem *davinci_intc_base;
> > +static struct irq_domain *davinci_irq_domain;
> >
> > static inline void davinci_irq_writel(unsigned long value, int offset)
> > {
> > @@ -55,17 +57,15 @@ static inline unsigned long davinci_irq_readl(int offset)
> > }
> >
> > static __init void
> > -davinci_alloc_gc(void __iomem *base, unsigned int irq_start, unsigned int num)
> > +davinci_irq_setup_gc(void __iomem *base,
> > + unsigned int irq_start, unsigned int num)
> > {
> > struct irq_chip_generic *gc;
> > struct irq_chip_type *ct;
> >
> > - gc = irq_alloc_generic_chip("AINTC", 1, irq_start, base, handle_edge_irq);
> > - if (!gc) {
> > - pr_err("%s: irq_alloc_generic_chip for IRQ %u failed\n",
> > - __func__, irq_start);
> > - return;
> > - }
> > + gc = irq_get_domain_generic_chip(davinci_irq_domain, irq_start);
>
> check for (gc == NULL) here?
>

I can add it, but it's not really needed. We know we pass correct
parameters to this routine and if it fails, the system won't boot
anyway.

Bart