On Fri, Mar 18, 2011 at 03:04:35PM -0700, David Miller wrote:
>
> Signed-off-by: David S. Miller <davem@xxxxxxxxxxxxx>
> ---
> arch/sparc/include/asm/irq_32.h | 3 +++
> arch/sparc/include/asm/irq_64.h | 2 ++
> 2 files changed, 5 insertions(+), 0 deletions(-)
>
> diff --git a/arch/sparc/include/asm/irq_32.h b/arch/sparc/include/asm/irq_32.h
> index cbf4801..eced3e3 100644
> --- a/arch/sparc/include/asm/irq_32.h
> +++ b/arch/sparc/include/asm/irq_32.h
> @@ -13,4 +13,7 @@
> #define irq_canonicalize(irq) (irq)
>
> extern void __init init_IRQ(void);
> +
> +#define NO_IRQ 0xffffffff
> +
> #endif
> diff --git a/arch/sparc/include/asm/irq_64.h b/arch/sparc/include/asm/irq_64.h
> index 4f09666f..16dcae6d 100644
> --- a/arch/sparc/include/asm/irq_64.h
> +++ b/arch/sparc/include/asm/irq_64.h
> @@ -97,4 +97,6 @@ extern void *softirq_stack[NR_CPUS];
> #define __ARCH_HAS_DO_SOFTIRQ
> #define ARCH_HAS_NMI_WATCHDOG
>
> +#define NO_IRQ 0xffffffff
> +
> #endif
Why -1 (0xffffffff)?
drivers/of/irq.c default to 0 - like this:
/* For archs that don't support NO_IRQ (such as x86), provide a dummy value */
#ifndef NO_IRQ
#define NO_IRQ 0
#endif
I can see that the actual value differs from arch to arch.
I recalled a discussion about this topic some years ago.
Do not ask how I managed to do so?!?
Anyway the input from Linus was to use 0 for NO_IRQ because
as he said:
====
But the point is that most drivers will do something like
if (!dev->irq)
return;
(whatever, made up). And that having NO_IRQ be anything but 0 is thus
fundamentally broken.
====
See the full mail here:
http://lkml.org/lkml/2005/11/21/211
I do not know of any places that would break - I just
wanted to bring this to your attention.
Sam
--
To unsubscribe from this list: send the line "unsubscribe sparclinux" in
the body of a message to majordomo@xxxxxxxxxxxxxxx
More majordomo info at http://vger.kernel.org/majordomo-info.html
