On Sat, May 02, 2009 at 01:34:03AM +0200, Atte André Jensen wrote:
> Dan S wrote:
> > Maybe your softwares support load-sharing over multiple processors (a
> > good thing, no?) and therefore all the work is being neatly divided
> > between the processors. Any reason that should be bad?
>
> No, what would be great. Unfortunately I'm pretty that's not what's
> happening. With the same patch I can run 35 (but not 40) voices in ams
> on dual core and 45 (but not 50) with one core disabled:
Just guessing, but it could be that with two cores
the audio thread (which would be responsible for most
of the load) gets scheduled randomly on one of the two
cores, which may have a penalty. This is possible as this
thread will sleep at the end of each period, waiting for
the next, and would also explain the nearly equal load
levels which you see.
Ciao,
-- FA Io lo dico sempre: l'Italia č troppo stretta e lunga. _______________________________________________ Linux-audio-user mailing list Linux-audio-user@email-addr-hidden http://lists.linuxaudio.org/mailman/listinfo/linux-audio-userReceived on Sat May 2 04:15:02 2009
This archive was generated by hypermail 2.1.8 : Sat May 02 2009 - 04:15:03 EEST