On 05/06/17 16:32, Boris Ostrovsky wrote:
I believe we do need to take affinity into consideration even if the
chance that it is non-default is small.
Agreed.
I am not opposed to having bind_last_selected_cpu percpu, I just wanted
to understand the reason better. Additional locking would be a downside
with a global so if you feel that percpu is worth it then I won't object.
If affinity == cpu_online_mask, then percpu will give a better spread.
atomic set/get can be used, if we want to use a global variable, but I
think it will be more random than percpu.
Yes, you are correct. .irq_set_affinity pretty much does the same thing.
The code will now looks like this.
raw_spin_lock_irqsave(lock, flags);
percpu read
select_cpu
percpu write
xen_rebind_evtchn_to_cpu(evtchn, selected_cpu)
raw_spin_unlock_irqsave(lock, flags);
(BTW, I just noticed --- you don't need to initialize desc)
Sorry, I didn't get it. Which desc doesn't need init ?
-Anoob
_______________________________________________
Xen-devel mailing list
Xen-devel@lists.xen.org
https://lists.xen.org/xen-devel
