Re: PREEMPT_RT vs ADEOS: the numbers, part 1

From: Kristian Benoit
Date: Sun Jun 12 2005 - 21:42:50 EST


On Sun, 2005-06-12 at 16:54 +0200, Ingo Molnar wrote:
> * James R Bruce <bruce@xxxxxxxxxxxxxx> wrote:
>
> > Ingo, if you could document the right options required for decent
> > performace somewhere it would be quite helpful (maybe in
> > Documentation/rt-preempt?). My first test of Preempt-RT showed
> > unexpectedly high overhead for a fairly benign network load (120 UDP
> > packets/sec), but that was likely the result of leaving some debugging
> > options on.
>
> agreed, this needs to be addressed.
>
> in the latest patch (-48-17 or later) i have changed the debugging
> options to default to off. (this wont turn them off if your .config has
> them turned on already, but will turn them off for new testers'
> .configs)
>
> I also added a prominent boot-time message that, if certain
> high-overhead debugging options are enabled, says:
>
> *****************************************************************************
> * *
> * WARNING, the following debugging options are turned on in your .config: *
> * *
> * CONFIG_DEBUG_RT_LOCKING_MODE *
> * CONFIG_RT_DEADLOCK_DETECT *
> * CONFIG_DEBUG_PREEMPT *
> * CONFIG_CRITICAL_PREEMPT_TIMING *
> * CONFIG_CRITICAL_IRQSOFF_TIMING *
> * CONFIG_LATENCY_TRACE *
> * CONFIG_DEBUG_SLAB *
> * *
> * they may increase runtime overhead and latencies considerably! *
> * *
> *****************************************************************************
>
> wrt. documentation - i'm not a big doc writer, but i'm taking patches
> :-)
>
> Ingo

Thanks, that will help us getting better results from PREEMPT_RT.

Kristian

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