Fusion and RT Rreempt [Was: RTAI/fusion 0.7.2]

Dmitry Adamushko dmitry.adamushko at datacon.at
Thu Jun 2 12:09:34 CEST 2005


rtai-admin at rtai.org wrote on 02.06.2005 11:03:58:

> Philippe Gerum wrote:
>  > Max Krasnyansky wrote:
>  > > This is totally reproduceable btw. I played with the BIOS settings
and
>  > > stuff,
>  > > makes no difference. MSI workaround is enabled but makes no
difference.
>
> SMI workaround is probably ineffective with this architecture, it
> currently works only with ICH based intel chipset. When SMI workaround
> detects a known chipset, it prints a clear message on the console. PIIX
> support maybe added soon. The problem with SMI workaround is that it is
> highly dependent on chipset.

We have 2 tests (as I understand, both have been conducted on the same
machine) and the difference is only a sampling period:

> 2.6.11, adeos r10, fusion CVS.
>
> latency -p 100 + high network load
> RTS|         343|       13153|       51667|       0|    00:02:30/00:02:30
>
> latency -p 4000 + high network load
> RTS|        1262|      780892|     1724248|       0|    00:00:19/00:00:19
>
> How about that :) ? 1.7 milliseconds. Wow.
>

The second one looks definitely like a result of SMI. At least I have
pretty same numbers on my P4 machine.
Then one thing I don't understand is why the first test shows completely
different numbers. The SMI must have taken effect in both cases, so why the
numbers are so different?
Have we got something like: the SMI depends on a sampling period?
I would expect to see the same latency caused by SMI for any sampling
period.

>
> --
>                    Gilles Chanteperdrix.

--
Dmitry
-------------- next part --------------
An HTML attachment was scrubbed...
URL: http://mail.rtai.org/pipermail/rtai/attachments/20050602/dfd3daa7/attachment.htm 


More information about the Rtai mailing list