Subject: Re: [linux-audio-dev] Re: costs of IPC
From: Abramo Bagnara (abramo_AT_alsa-project.org)
Date: Wed May 16 2001 - 14:22:07 EEST
Steve Harris wrote:
>
> On Tue, May 15, 2001 at 11:06:20PM +0200, Abramo Bagnara wrote:
> > I'd like you plot the difference and difference/workers for workers = 4
> > and workers = 8. If I'm not wrong the difference/workers should never be
> > greater than about 2.5%.
>
> http://plugin.org.uk/spmp/worker-diff.png
>
> The y axis is the difference (per worker) in CPU use % between the sp and
> mp case. The x axis the size of the worker polution in k. Shared pollution
> was 4k and stack 1k.
>
> I wouldn't like to draw any conclusions from that myself!
>
SMP?
as root?
Could you publish your scripts then we can try the very same tests on
different machines? It would be wonderful if graph generation is
somewhat automatic from script run, so no skill is needed (apart
parameter setting).
That apart, my current conclusion is that multiprocess approach is
definitely interesting and we need to design an architecture that works
fine with both approaches. Period.
All the rest is very funny and terribly interesting, but this conclusion
is not opinable IMNSHO.
Do you agree?
-- Abramo Bagnara mailto:abramo_AT_alsa-project.orgOpera Unica Phone: +39.546.656023 Via Emilia Interna, 140 48014 Castel Bolognese (RA) - Italy
ALSA project http://www.alsa-project.org It sounds good!
This archive was generated by hypermail 2b28 : Wed May 16 2001 - 14:43:08 EEST