aplay lists both Playback and Capture devices

aplay lists both Playback and Capture devices

[Fabio Estevam]

Hi,

When running 'aplay -l' and 'arecord -l' I see the following output on
a mx28evk running a 3.10-rc kernel:

/$ aplay -l
**** List of PLAYBACK Hardware Devices ****
card 0: mxssgtl5000 [mxs_sgtl5000], device 0: Playback sgtl5000-0 []
  Subdevices: 1/1
  Subdevice #0: subdevice #0
card 0: mxssgtl5000 [mxs_sgtl5000], device 1: Capture sgtl5000-1 []
  Subdevices: 1/1
  Subdevice #0: subdevice #0

$ arecord -l
**** List of CAPTURE Hardware Devices ****
card 0: mxssgtl5000 [mxs_sgtl5000], device 0: Playback sgtl5000-0 []
  Subdevices: 1/1
  Subdevice #0: subdevice #0
card 0: mxssgtl5000 [mxs_sgtl5000], device 1: Capture sgtl5000-1 []
  Subdevices: 1/1
  Subdevice #0: subdevice #0

I would expect to see only the Playback entry in aplay and only the
Record entry in arecord.

Is there anything I should change in sound/soc/mxs/mxs-sgtl5000.c in
order to get the correct outputs?

Regards,

Fabio Estevam

Re: aplay lists both Playback and Capture devices

[Mark Brown]

[-- Attachment #1.1: Type: text/plain, Size: 195 bytes --]

On Wed, May 29, 2013 at 06:32:56PM -0300, Fabio Estevam wrote:

> Is there anything I should change in sound/soc/mxs/mxs-sgtl5000.c in
> order to get the correct outputs?

No, this is userspace.

[-- Attachment #1.2: Digital signature --]
[-- Type: application/pgp-signature, Size: 836 bytes --]

[-- Attachment #2: Type: text/plain, Size: 0 bytes --]

Re: aplay lists both Playback and Capture devices

[Clemens Ladisch]

Fabio Estevam wrote:
> When running 'aplay -l' and 'arecord -l' I see the following output on
> a mx28evk running a 3.10-rc kernel:
>
> $ aplay -l
> card 0: mxssgtl5000 [mxs_sgtl5000], device 0: Playback sgtl5000-0 []
> card 0: mxssgtl5000 [mxs_sgtl5000], device 1: Capture sgtl5000-1 []
> $ arecord -l
> card 0: mxssgtl5000 [mxs_sgtl5000], device 0: Playback sgtl5000-0 []
> card 0: mxssgtl5000 [mxs_sgtl5000], device 1: Capture sgtl5000-1 []
>
> I would expect to see only the Playback entry in aplay and only the
> Record entry in arecord.

Please show the contents of /proc/asound/devices.


Regards,
Clemens

Re: aplay lists both Playback and Capture devices

[Fabio Estevam]

Hi Clemens,

On Thu, May 30, 2013 at 8:23 AM, Clemens Ladisch <clemens@ladisch.de> wrote:

>> I would expect to see only the Playback entry in aplay and only the
>> Record entry in arecord.
>
> Please show the contents of /proc/asound/devices.

Here it goes:

root@freescale /$ cat /proc/asound/devices
  0: [ 0]   : control
 16: [ 0- 0]: digital audio playback
 17: [ 0- 1]: digital audio playback
 24: [ 0- 0]: digital audio capture
 25: [ 0- 1]: digital audio capture
 33:        : timer

Thanks,

Fabio Estevam

Re: aplay lists both Playback and Capture devices

[Fabio Estevam]

On Sat, Jun 1, 2013 at 3:02 PM, Fabio Estevam <festevam@gmail.com> wrote:
> Hi Clemens,
>
> On Thu, May 30, 2013 at 8:23 AM, Clemens Ladisch <clemens@ladisch.de> wrote:
>
>>> I would expect to see only the Playback entry in aplay and only the
>>> Record entry in arecord.
>>
>> Please show the contents of /proc/asound/devices.
>
> Here it goes:
>
> root@freescale /$ cat /proc/asound/devices
>   0: [ 0]   : control
>  16: [ 0- 0]: digital audio playback
>  17: [ 0- 1]: digital audio playback
>  24: [ 0- 0]: digital audio capture
>  25: [ 0- 1]: digital audio capture
>  33:        : timer

Also, in case it helps:

root@freescale /$ cat /proc/asound/pcm
00-00: HiFi Playback sgtl5000-0 :  : playback 1 : capture 1
00-01: HiFi Capture sgtl5000-1 :  : playback 1 : capture 1

root@freescale /$ cat /proc/asound/
card0/        devices       pcm           version
cards         mxssgtl5000/  timers

root@freescale /$ cat /proc/asound/cards
 0 [mxssgtl5000    ]: mxs_sgtl5000 - mxs_sgtl5000
                      mxs_sgtl5000

root@freescale /$ cat /proc/asound/card0/
id      pcm0c/  pcm0p/  pcm1c/  pcm1p/

root@freescale /$ cat /proc/asound/card0/pcm0c/info
card: 0
device: 0
subdevice: 0
stream: CAPTURE
id: HiFi Playback sgtl5000-0
name:
subname: subdevice #0
class: 0
subclass: 0
subdevices_count: 1
subdevices_avail: 1

root@freescale /$ cat /proc/asound/card0/pcm0p/info
card: 0
device: 0
subdevice: 0
stream: PLAYBACK
id: HiFi Playback sgtl5000-0
name:
subname: subdevice #0
class: 0
subclass: 0
subdevices_count: 1
subdevices_avail: 1
root@freescale /$

Regards,

Fabio Estevam

Re: aplay lists both Playback and Capture devices

[Mark Brown]

[-- Attachment #1.1: Type: text/plain, Size: 365 bytes --]

On Sat, Jun 01, 2013 at 03:09:38PM -0300, Fabio Estevam wrote:

> root@freescale /$ cat /proc/asound/pcm
> 00-00: HiFi Playback sgtl5000-0 :  : playback 1 : capture 1
> 00-01: HiFi Capture sgtl5000-1 :  : playback 1 : capture 1

Hrm, in that case the kernel does seem to be lying about capabilites...
are all the DAIs correctly listing that they're unidirectional?

[-- Attachment #1.2: Digital signature --]
[-- Type: application/pgp-signature, Size: 836 bytes --]

[-- Attachment #2: Type: text/plain, Size: 0 bytes --]

Re: aplay lists both Playback and Capture devices

[Fabio Estevam]

Hi Mark,

On Sat, Jun 1, 2013 at 4:06 PM, Mark Brown <broonie@kernel.org> wrote:
> On Sat, Jun 01, 2013 at 03:09:38PM -0300, Fabio Estevam wrote:
>
>> root@freescale /$ cat /proc/asound/pcm
>> 00-00: HiFi Playback sgtl5000-0 :  : playback 1 : capture 1
>> 00-01: HiFi Capture sgtl5000-1 :  : playback 1 : capture 1
>
> Hrm, in that case the kernel does seem to be lying about capabilites...
> are all the DAIs correctly listing that they're unidirectional?

What is the proper way to know this information?

Or how do I tell that each DAI is unidirectional?

Thanks,

Fabio Estevam

Re: aplay lists both Playback and Capture devices

[Mark Brown]

[-- Attachment #1.1: Type: text/plain, Size: 223 bytes --]

On Sat, Jun 01, 2013 at 04:20:50PM -0300, Fabio Estevam wrote:

> What is the proper way to know this information?

> Or how do I tell that each DAI is unidirectional?

If it's not got any playback or capture capabilities.

[-- Attachment #1.2: Digital signature --]
[-- Type: application/pgp-signature, Size: 836 bytes --]

[-- Attachment #2: Type: text/plain, Size: 0 bytes --]

Re: aplay lists both Playback and Capture devices

[Fabio Estevam]

On Sat, Jun 1, 2013 at 6:05 PM, Mark Brown <broonie@kernel.org> wrote:
> On Sat, Jun 01, 2013 at 04:20:50PM -0300, Fabio Estevam wrote:
>
>> What is the proper way to know this information?
>
>> Or how do I tell that each DAI is unidirectional?
>
> If it's not got any playback or capture capabilities.

On mx28evk there are two DAIs connected to a single sgtl5000:

- One corresponds to the serial audio interface 0 (SAIF0) and it is
playback only

- The other corresponds to the serial audio interface 1 (SAIF1) and it
is capture only.

Should I create and pass device tree bindings like "playback-only" and
"capture-only", so that we can avoid having both playback and capture
capabilities for SAIF0 and SAIF1?

Or is there already a mechanism in place that I can use to configure a
DAI as unidirectional?

Thanks,

Fabio Estevam

Re: aplay lists both Playback and Capture devices

[Mark Brown]

[-- Attachment #1.1: Type: text/plain, Size: 1267 bytes --]

On Sat, Jun 01, 2013 at 06:36:49PM -0300, Fabio Estevam wrote:
> On Sat, Jun 1, 2013 at 6:05 PM, Mark Brown <broonie@kernel.org> wrote:

> > If it's not got any playback or capture capabilities.

> On mx28evk there are two DAIs connected to a single sgtl5000:

> - One corresponds to the serial audio interface 0 (SAIF0) and it is
> playback only

> - The other corresponds to the serial audio interface 1 (SAIF1) and it
> is capture only.

These are not DAIs, they are DAI links.  The capabilities of the links
should be being worked out by examining

> Should I create and pass device tree bindings like "playback-only" and
> "capture-only", so that we can avoid having both playback and capture
> capabilities for SAIF0 and SAIF1?

No, it's not sensible to require users to set DT properties for basic
features of the harwdare.  If for some reason the interface could be
bidirectional but the board design breaks that for some reason then it
may make sense but otherwise the kernel ought to be able to work this
out based on the fact that one or both ends of the link can only do one
direction.

> Or is there already a mechanism in place that I can use to configure a
> DAI as unidirectional?

How is the kernel currently managing to get properties for playback?

[-- Attachment #1.2: Digital signature --]
[-- Type: application/pgp-signature, Size: 836 bytes --]

[-- Attachment #2: Type: text/plain, Size: 0 bytes --]

Re: aplay lists both Playback and Capture devices

[Marek Vasut]

Dear Mark Brown,

> On Sat, Jun 01, 2013 at 04:20:50PM -0300, Fabio Estevam wrote:
> > What is the proper way to know this information?
> > 
> > Or how do I tell that each DAI is unidirectional?
> 
> If it's not got any playback or capture capabilities.

Each of the SAIFs has only one direction, the question Fabio probably wanted to 
ask is where he can fix the SAIF caps so they report only one direction instead 
of both.

Best regards,
Marek Vasut

Re: aplay lists both Playback and Capture devices

[Mark Brown]

[-- Attachment #1.1: Type: text/plain, Size: 455 bytes --]

On Sat, Jun 01, 2013 at 11:42:21PM +0200, Marek Vasut wrote:

> Each of the SAIFs has only one direction, the question Fabio probably wanted to 
> ask is where he can fix the SAIF caps so they report only one direction instead 
> of both.

Only specifying playback or capture capabilities ought to do the job.
Looking at soc-pcm.c it rather looks like things got broken during the
soc-pcm stuff if they ever worked for CPU side limitations here.

[-- Attachment #1.2: Digital signature --]
[-- Type: application/pgp-signature, Size: 836 bytes --]

[-- Attachment #2: Type: text/plain, Size: 0 bytes --]