Re: rt20 scheduling latency testcase and failure data

[Darren Hart <dvhltc@us.ibm.com>]

On Monday 15 May 2006 04:15, Sébastien Dugué wrote:

>   I've been running you test program on my box under a stress-kernel
> load and did not observe any failure as you describe, not even a max
> latency overshooting the 100 us limit (max latencies in the 60~70 us).
>
>   I even went to decrease PERIOD to 1 ms and still no failure.
>
>   I'm running rt20 with the futex priority based wakeup patch on
> a dual 2.8 GHz HT Xeon box. All hardirq and softirq threads are at their
> default priority.

Interesting, I'll have to try this on some more hardware and see if I can 
reproduce there.

>
>   How do you generate the network load you mention? Maybe I could try at
> least with the same load you're using.

I was simply copying a 60MB file to the test machine via scp, in a bash while 
loop.  I haven't been doing this on my most recent runs however, and they 
still fail.  So I don't believe the net load is directly related.

I am going to work with Ingo's trace-it.c today and report back.

-- 
Darren Hart
IBM Linux Technology Center
Realtime Linux Team