Re: [PATCH Dovetail v2 1/2] dovetail: genirq: Add request_percpu_irq_affinity_flags()

Florian Bezdeka <[email protected]> writes:

> On Wed, 2026-02-18 at 08:30 +0100, Jan Kiszka wrote:
>> On 18.02.26 00:05, Florian Bezdeka wrote:
>> > On Tue, 2026-02-17 at 23:54 +0100, Florian Bezdeka wrote:
>> > > cobalt / evl are currently using __request_percpu_irq() for setting up
>> > > OOB IRQs. In 6.19 the signature of __request_percpu_irq() has been
>> > > updated and 7.0 will remove the function entirely, so we need a
>> > > replacement.
>> > > 
>> > > All other variants (like the upcoming request_percpu_irq_affinity())
>> > > are missing the flags parameter which we require.
>> > > 
>> > > This could be backported to all supported dovetail versions. That
>> > > would allow co-kernels to switch to the new API unconditionally.
>> > > 
>> > > Signed-off-by: Florian Bezdeka <[email protected]>
>> > > ---
>> > >  include/linux/interrupt.h |  5 +++++
>> > >  kernel/irq/manage.c       | 10 ++++++++++
>> > >  2 files changed, 15 insertions(+)
>> > > 
>> > > diff --git a/include/linux/interrupt.h b/include/linux/interrupt.h
>> > > index bf0511b438c65e119af917763ce92a8f3bfb4c4c..6cd8ad02ffbd87f71869d6737fbc8c1ae477c1ee 100644
>> > > --- a/include/linux/interrupt.h
>> > > +++ b/include/linux/interrupt.h
>> > > @@ -213,6 +213,11 @@ request_percpu_irq_affinity(unsigned int irq, irq_handler_t handler,
>> > >  				    devname, affinity, percpu_dev_id);
>> > >  }
>> > >  
>> > > +int request_percpu_irq_affinity_flags(unsigned int irq, irq_handler_t handler,
>> > > +				      unsigned long flags, const char *devname,
>> > > +				      const cpumask_t *affinity,
>> > > +				      void __percpu *dev_id);
>> > > +
>> > > 
>> > 
>> > Would be nice to get some feedback here. The affinity parameter is
>> > currently not in use, but would allow us to provide it in the future.
>> > Not sure if that will be necessary...
>> > 
>> 
>> This is what I was wondering as well... When would one need that? Maybe
>> in the context of confining Xenomai to a subset of physical CPUs? Would
>> be good to check that again.
>
> @Philippe: Any thoughts?
>
> I would like to complete the 6.19 port on Xenomai side. This is a
> blocker right now.
>

To me this is a no-brainer, given that __request_percpu_irq() requires
an affinity mask, having request_percpu_irq_affinity_flags() is the only
sane way to set all the properties of a per-cpu interrupt, including
flags and affinity, in one go. In any case, callers who don't care about
such affinity can pass NULL.

-- 
Philippe.