From: Rafaël Carré <rafael.carre_at_gmail.com>
Date: Fri, 7 May 2010 23:35:39 +0200

On Fri, 7 May 2010 23:07:46 +0200
Marcin Bukat <marcin.bukat_at_gmail.com> wrote:

> I am trying to find out why playback on MPIO HD200 needs much more cpu
> than on other coldfire based DAPs.
> 1) test_codec gives comparable results to other CFs.
> 2) display drivers are pretty fast scoring ~764fps fullscreen updates
> in test_fps.
> 3) test_mem gives on MPIO: unboosted write 95kB/s, ubusted read 55
> kB/s, boosted write 267kB/s and boosted read 154kB/s. Pixelma posted
> some results on IRC from iaudio M5 "bufsize: 514048, write ticks: 15
> (267 kB/s), read ticks: 26 (154 kB/s) - no change when boosted or
> unboosted". So here is big difference when unboosted.
> 4) disk transfers are comparable to CompactFlash modded ipod mini (I
> am testing with CompactFlash also)
> 5) playing wav MPIO runs unboosted after initial rebuffer.

You've tested a lot of things already so hopefully you're not far now.

> So basicaly I would like to ask for ideas where may be the problem. I
> would like to ask also for test_mem results from other CF devices. For
> me it looks a bit strange that Pixelma doesn't see difference when
> boosted/unboosted.

Did you try to run a profiled build ?

Is there a fast counter which could measure the time taken by each tick?

-- 
Rafaël Carré

Received on 2010-05-07