* DAPM widget power off
@ 2013-04-18 7:02 yuvaraj r
2013-04-20 17:19 ` Gabriel M. Beddingfield
0 siblings, 1 reply; 2+ messages in thread
From: yuvaraj r @ 2013-04-18 7:02 UTC (permalink / raw)
To: alsa-devel
Hi All,
I have two different dais
1. Hifi Multimedia
2. Voice call dai
Before opening voice call dai. I closed Hifi stream. After 5 sec Voice call
stream dapm widget got powered off automatically. I am assuming that
stream close event of Multimedia causing problem.
What will issue in my cause.
Thanks
Yuvaraj
^ permalink raw reply [flat|nested] 2+ messages in thread
* Re: DAPM widget power off
2013-04-18 7:02 DAPM widget power off yuvaraj r
@ 2013-04-20 17:19 ` Gabriel M. Beddingfield
0 siblings, 0 replies; 2+ messages in thread
From: Gabriel M. Beddingfield @ 2013-04-20 17:19 UTC (permalink / raw)
To: yuvaraj r; +Cc: alsa-devel
On 04/18/2013 12:02 AM, yuvaraj r wrote:
> Hi All,
>
> I have two different dais
>
> 1. Hifi Multimedia
>
> 2. Voice call dai
>
> Before opening voice call dai. I closed Hifi stream. After 5 sec Voice call
> stream dapm widget got powered off automatically. I am assuming that
> stream close event of Multimedia causing problem.
>
> What will issue in my cause.
The value of snd_soc_card::pmdown_time and
snd_soc_pcm_runtime::pmdown_time are initialized to 5000 (i.e. 5 secs)
by default. That means that after you close a stream, the kernel waits
5 seconds and then powers everything down.
There's a few reasons why this can happen:
1. Your DAPM routing doesn't adequately cover the Voice call case.
Something that is needed for both Voice and Multimedia isn't
properly associated with the Voice routing... so DAPM shuts
it down because it's not "in use."
2. Some of your power-down code is making a poor assumption. I.e.
you're making some decisions outside of DAPM about when a
component should be powered down, and it's the wrong decision.
3. I've seen a bug in older kernels (e.g. <= 3.1) where the DAPM
logic will accidentally shut down a capture stream while powering
down a playback stream. (E.g. see
http://review.omapzoom.org/#/c/21405/)
A quick work-around is to set pmdown_time = 0. However, there will
still be a race condition... so don't consider it "fixed." (Though, I'm
pretty sure I've seen a device ship with this "fix." :-))
-gabriel
^ permalink raw reply [flat|nested] 2+ messages in thread
end of thread, other threads:[~2013-04-20 17:19 UTC | newest]
Thread overview: 2+ messages (download: mbox.gz follow: Atom feed
-- links below jump to the message on this page --
2013-04-18 7:02 DAPM widget power off yuvaraj r
2013-04-20 17:19 ` Gabriel M. Beddingfield
This is an external index of several public inboxes,
see mirroring instructions on how to clone and mirror
all data and code used by this external index.