Peter,

pet...@infradead.org writes:
> On Thu, Aug 06, 2020 at 11:41:06AM +0200, Thomas Gleixner wrote:
>> And that has nothing to do
>> with a silly test case. Sporadic runaways due to a bug in a once per
>> week code path simply can happen and having the safety net working
>> depending on a config option selected or not is just wrong.
>
> The safety thing is concerned with RT tasks. It doesn't pretend to help
> with runnaway IRQs, never has, never will.

Of course not. But without irq accounting the runtime is accounted on
the runaway task which causes it to throttle.

> The further extreme is an interrupt storm, those have always taken a
> machine down.

If every interrupt is actually handled, then yes.

> Accounting unrelated IRQ time to RT tasks is equally wrong, the task
> execution is unrelated to the IRQs. The config option at least offers
> insight into where time goes -- and it's a config option because doing
> time accounting on interrupts adds overhead :/

Right, but it's not totally out of the world either to make the
throttler do:

          if (rt_runtime + irq_time > threshold)
                try_to_keep_the_box_alive()

> This really is a no-win all round.

That's not the question here :)

> The only 'sensible' option here is threaded IRQs, where the IRQ line
> gets disabled until the handler thread has ran, that also helps with IRQ
> storms.

I'm not against enforcing threaded IRQs. :)

Thanks,

        tglx

Reply via email to