snd_soc_jack_add_gpios vs. SOC_DAPM_PIN_SWITCH

snd_soc_jack_add_gpios vs. SOC_DAPM_PIN_SWITCH

[Stephen Warren]

I've set up headphone hack detection in my machine driver, using
snd_soc_jack_new, snd_soc_jack_add_pins, and snd_soc_jack_add_gpios.

I've also created a control to allow the user to enable/disable the headphone
jack:

static const struct snd_kcontrol_new harmony_controls[] = {
	SOC_DAPM_PIN_SWITCH("Headphone Jack"),
};

On their own, both of these work fine.

However, if I use the control to disable the headphones, then unplug/replug
the headphones, the jack detect seems to override the control, and I hear
audio from the headphones.

Is this something that's known not yet implemented or was never intended to
be used this way, or should I look further to find the issue?

FYI, sound/soc/samsung/smartq_wm8987.c has this same code structure, and I 
imagine will have the same issue.

Thanks.

-- 
nvpublic

Re: snd_soc_jack_add_gpios vs. SOC_DAPM_PIN_SWITCH

[Mark Brown]

On Mon, Jan 31, 2011 at 04:03:58PM -0800, Stephen Warren wrote:

> However, if I use the control to disable the headphones, then unplug/replug
> the headphones, the jack detect seems to override the control, and I hear
> audio from the headphones.

> Is this something that's known not yet implemented or was never intended to
> be used this way, or should I look further to find the issue?

Using these two together makes no sense.  Either you manually control
the state of the output or you manage it automatically.

Generally a PIN_SWITCH() should only be used to offer control if none is
availabile within the CODEC - in the case of Harmony it wouldn't make
much sense as there's plenty of control in the CODEC for what gets
routed to the various outputs.

> FYI, sound/soc/samsung/smartq_wm8987.c has this same code structure, and I 
> imagine will have the same issue.

Probably.

Re: snd_soc_jack_add_gpios vs. SOC_DAPM_PIN_SWITCH

[Stephen Warren]

Mark Brown wrote at Tuesday, February 01, 2011 4:34 AM:
> On Mon, Jan 31, 2011 at 04:03:58PM -0800, Stephen Warren wrote:
> 
> > However, if I use the control to disable the headphones, then
> > unplug/replug the headphones, the jack detect seems to override
> > the control, and I hear audio from the headphones.
> 
> Using these two together makes no sense.  Either you manually control
> the state of the output or you manage it automatically.

I was thinking that the jack detect would be used to enable/disable widgets
for power purposes, but with a manual override to allow SW to set up use
cases where only the internal speakers were used, and never headphones.

That said, if that isn't something that's ever typically done, I'm fine
not supporting that.

> Generally a PIN_SWITCH() should only be used to offer control if none is
> availabile within the CODEC - in the case of Harmony it wouldn't make
> much sense as there's plenty of control in the CODEC for what gets
> routed to the various outputs.

IIRC, the codec controls allow selecting only what gets routed, but there
wasn't an overall "enable/disable speakers" switch (the mute switch does
not disable the DAPM widgets). One would have to disable all the routing
options to disable the speaker amplifiers, which would entail remembering
what those routing options were to restore them once the speakers were
desired again. A manual switch avoids this.

-- 
nvpublic

Re: snd_soc_jack_add_gpios vs. SOC_DAPM_PIN_SWITCH

[Mark Brown]

On Tue, Feb 01, 2011 at 10:59:00AM -0800, Stephen Warren wrote:
> Mark Brown wrote at Tuesday, February 01, 2011 4:34 AM:

> > Using these two together makes no sense.  Either you manually control
> > the state of the output or you manage it automatically.

> I was thinking that the jack detect would be used to enable/disable widgets
> for power purposes, but with a manual override to allow SW to set up use
> cases where only the internal speakers were used, and never headphones.

> That said, if that isn't something that's ever typically done, I'm fine
> not supporting that.

It's really uncommon - generally if headphones are inserted you want to
play audio on them even if also using the speakers as when people have
headphones in they may not be able to hear the speakers.  Headphones are
often pretty good earplugs too.

> > Generally a PIN_SWITCH() should only be used to offer control if none is
> > availabile within the CODEC - in the case of Harmony it wouldn't make
> > much sense as there's plenty of control in the CODEC for what gets
> > routed to the various outputs.

> IIRC, the codec controls allow selecting only what gets routed, but there
> wasn't an overall "enable/disable speakers" switch (the mute switch does
> not disable the DAPM widgets). One would have to disable all the routing
> options to disable the speaker amplifiers, which would entail remembering
> what those routing options were to restore them once the speakers were
> desired again. A manual switch avoids this.

It may make sense to have a pin switch for the speakers, it's much more
of a userspace question if you want the speakers on when the headphones
are inserted so they probably shouldn't be connected to the jack at all
(as we discussed previously).  Bear in mind that in the *very* common
case where the outputs are connected to DACs only everything will get
powered off when there's no audio playing anyway.

Having looked at the WM8903 again this probably does make most sense for
the speakers.