Thanks Dave, it worked fine. We're using an own implementation for ALSA, but
I think that also alsa_sound implementation will have the same problem.
We'll move to that implementation soon, so I think it's a good idea to start
working on it.

-Misa
2009/1/23 Dave Sparks <[email protected]>

>
> The correct place to spin up the audio hardware is in the
> AudioOutputStream::write() method. In our other implementations, we
> have a standby flag that is checked at the start of write and if true,
> the kernel driver is opened. In AudioOutputStream::standby(), we close
> the driver and set the standby flag to true.
>
> On Jan 23, 7:24 pm, Misael Lopez <[email protected]> wrote:
> > We are having problems trying to put system in "suspend" state because
> some
> > clock are kept alive. These clocks are audio related.
> >
> > AudioFlinger's constructor calls openOutputStream method, which opens
> ALSA
> > pcm interface by calling "snd_pcm_open" (ALSA API). In ALSA kernel driver
> it
> > means that resources will be allocated and hardware codec will be on.
> > However, even if no output stream is being transmitted, stream channel is
> > never closed making audio resources being active always at kernel level.
> > This prevents the system to sleep.
> >
> > I think we should only open ALSA PCM interface when needed and not
> directly
> > from boot-time, but it seems that openOutputStream method is the right
> place
> > to call "snd_pcm_open", isn't it? Is there any other better place to open
> > ALSA interface? Or any other solution?
> >
> > -Misa
> >
>

--~--~---------~--~----~------------~-------~--~----~
You received this message because you are subscribed to the Google Groups 
"android-framework" group.
To post to this group, send email to [email protected]
To unsubscribe from this group, send email to 
[email protected]
For more options, visit this group at 
http://groups.google.com/group/android-framework?hl=en
-~----------~----~----~----~------~----~------~--~---

Reply via email to