Re: crypto: caam from tasklet to threadirq

[Thomas Gleixner <tglx@linutronix.de>]

On Tue, 20 Sep 2016, Russell King - ARM Linux wrote:
> which corresponds to an 8% slowdown for the threaded IRQ case.  So,
> tasklets are indeed faster than threaded IRQs.

Fair enough.

> I've tried to perf it, but...
>  ....
 
> So, sorry, I'm not going to bother trying to get any further with this.
> If the job was not made harder stupid hardware design and kernel
> politics, then I might be more inclined to do deeper investigation, but
> right now I'm finding that I'm not interested in trying to jump through
> these stupid hoops.

I'd be very interested in a sched_switch + irq + softirq trace which does
not involve PMU hardware for both irqthreads and tasklets, but I can
understand if you can't be bothered to gather it.

Vs. the PMU interrupt thing. What's the politics about that? Do you have
any pointers?
 
> I think I've proven from the above that this patch needs to be reverted
> due to the performance regression, and that there _is_ most definitely
> a deterimental effect of switching from tasklets to threaded IRQs.

I agree that the revert should happen, but I'd rather see a bit more
information why this regression happens with the switch from tasklets to
threaded irqs.

Thanks,

	tglx