Re: Re: [PATCH] emu10k1: add interval timer support

[Takashi Iwai <tiwai@suse.de>]

At Fri, 24 Sep 2004 11:26:15 -0400,
Paul Davis wrote:
> 
> >> more to the point, perhaps, is that the very first implementation of
> >> the RME Hammerfall (digi9652) driver that i did took lee's approach
> >> and when we measured it, there was a very definite performance
> >> hit. having to call the "elapsed" function for every mono stream added
> >> measurable overhead. when we collapsed it back to a single
> >> non-interleaved, 26 channel device, it was significantly more
> >> efficient.
> >
> >It's a good point.  If you use the timer-style irq, the perfomance
> >decrease would be smaller, though, because you don't (can't) get irqs
> >per PCM stream.  However, the overhead must still exist.
> 
> it was the same with the digi9652 - one interrupt source per card
> card, but the elapsed callback had to be made to each stream.

OK, then is it the overhead of kernel/user switches?
If we use multi mono streams, poll is called N times.  I can imagine
this would hit performance loss.

Other possibility is the overhead of the task done in
snd_pcm_period_elapsed().  But if this really hits, we have to
eliminate it.


Takashi


-------------------------------------------------------
This SF.Net email is sponsored by: YOU BE THE JUDGE. Be one of 170
Project Admins to receive an Apple iPod Mini FREE for your judgement on
who ports your project to Linux PPC the best. Sponsored by IBM.
Deadline: Sept. 24. Go here: http://sf.net/ppc_contest.php