Re: [PATCH 4/5] Centralise NO_IRQ definition

[David Woodhouse <dwmw2@infradead.org>]

On Tue, 2005-11-22 at 18:37 +0000, David Howells wrote:
> 
>  (3) Having to translate a cookie for a specific IRQ means that the IRQ
>      handling code will be slower and more complex, or is going to avoid the
>      issue and be naughty and not deal with irq == NO_IRQ properly:
> 
>      The x86 PIC reports it as IRQ 0 having happened. In which case, by your
>      argument, you _have_ to translate it: you're not allowed to pass NO_IRQ to
>      setup_irq(), and you're not allowed to pass it to the interrupt handler -
>      in this case timer_interrupt(). Doing otherwise is wrong, insane, etc...

This is true. If we're suddenly going to start pretending that IRQ 0
isn't a valid interrupt merely on the basis that "x86 doesn't use it"¹,
then we can't really go making an exception to allow us to use IRQ 0 on
i386.

-- 
dwmw2
¹ ...despite the fact that even that isn't true on legacy-free machines.

-
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo@vger.kernel.org
More majordomo info at  http://vger.kernel.org/majordomo-info.html
Please read the FAQ at  http://www.tux.org/lkml/