Heiko Carstens wrote:
quoted text
> Hmm.. probably a stupid question: but what could happen that a real cpu
> (not virtual) becomes unresponsive so that it won't schedule a MAX_RT_PRIO-1
> prioritized task for 5 seconds?

The original problem (once I heard and easily reproduced) was there was an
another MAX_RT_PRIO-1 task and the task was spinning in itself by a bug.
(Now this would not be a problem since RLIMIT_RTTIME will work for it, but
 I cannot deny that there are some situations which cannot set the limit.)

However there would be more possible problem in the world, ex. assume that
a routine work with interrupt (and also preemption) disabled have an issue
of scalability so it takes long time on huge machine then stop_machine will
stop whole system such long time.  You can assume a driver's bug.  Now the
stop_machine is good tool to escalate a partial problem to global suddenly.

quoted text
>> So I think monotonic wallclock time actually makes the most sense here.
> 
> This is asking for trouble... a config option to disable this would be
> nice. But as I don't know which problem this patch originally addresses
> it might be that this is needed anyway. So lets see why we need it first.

I'm not good at VM etc., but I think user doesn't care who holds a cpu,
whether other guest or actual buggy software or space alien or so.
The important thing here is return control to user if timeout.

Thanks,
H.Seto
--
unsubscribe notice
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo@vger.kernel.org
More majordomo info at  http://vger.kernel.org/majordomo-info.html
Please read the FAQ at  http://www.tux.org/lkml/