On Fri, Oct 16, 2020 at 04:41:50PM +0200, Heiner Kallweit wrote:
> On 16.10.2020 16:26, Vladimir Oltean wrote:
> > On Fri, Oct 16, 2020 at 01:34:55PM +0200, Heiner Kallweit wrote:
> >> I'm aware of the topic, but missing the benefits of the irqoff version
> >> unconditionally doesn't seem to be the best option.
> >
> > What are the benefits of the irqoff version? As far as I see it, the
> > only use case for that function is when the caller has _explicitly_
> > disabled interrupts.
> >
> If the irqoff version wouldn't have a benefit, then I think we wouldn't
> have it ..
>
> > The plain napi_schedule call will check if irqs are disabled. If they
> > are, it won't do anything further in that area. There is no performance
> > impact except for a check.
> >
> There is no such check, and in general currently attempts are made to
> remove usage of e.g. in_interrupt(). napi_schedule() has additional calls
> to local_irq_save() and local_irq_restore().
This has nothing to do with in_interrupt().
Now, to explain where my confusion came from.
arm64 has this:
static inline unsigned long arch_local_irq_save(void)
{
unsigned long flags;
flags = arch_local_save_flags();
/*
* There are too many states with IRQs disabled, just keep the current
* state if interrupts are already disabled/masked.
*/
if (!arch_irqs_disabled_flags(flags))
arch_local_irq_disable();
return flags;
}
I just thought that the generic implementation had the "if" too.
