On Sun, 9 Oct 2016, Rich Felker wrote:
> Ideas for improvement are welcome -- for example the
> irq_is_percpu(irq_desc_get_irq(desc)) thing looks rather silly but I
See the other mail.
> didn't see a better way without poking through abstractions -- but
> overall I think this both solves the timer stall issue that I wasted
> other people's time on, and addresses the concerns about the J-Core
> AIC driver being oblivious to whether an irq is per-cpu.
You could put that knowledge into the device tree so you can decide at
mapping time whether it is per cpu or not.
Thanks,
tglx
