Native preemption is the culprit [was Re: today's CURRENT lockups]

[Martin Blapp]

Hi,

> I'm now experiencing extremely hard hangs in the following configurations:
>
>   SMP kernel running SCHED_ULE with hyperthreads
>   SMP kernel running SCHED_4BSD with hyperthreads
>
> To generate the load, I'm using the "supersmack" benchark with the
> select-key.smack query set with 30 client workers and 10,000 transactions.
> I am able to reliable hang the system with one or two runs.

I'm seeing the same effect here on a 2 CPU IBM server running without HTT. The
server is completly locked. We have running mimedefang on it and clamd, after
2-3 hours running, suddenly the processlist grows and grows and the box is not
accessibe anymore.

> By disabling the "#define PREEMPTION" entry in param.h with SCHED_4BSD,
> I'm able to complete the benchmark several times in a row without apparent
> problems. However, I'll leave it running for a few more hours and see if I
> didn't just "get lucky".  I'll then try SCHED_ULE w/o PREEMPTION.

I'll try that too.

Martin