mbox series

[00/17] ASoC: remove snd_soc_dai_set_tristate()

Message ID 87wmg9mhkk.wl-kuninori.morimoto.gx@renesas.com (mailing list archive)
Headers show
Series ASoC: remove snd_soc_dai_set_tristate() | expand

Message

Kuninori Morimoto Dec. 9, 2024, 7:38 a.m. UTC
Hi Mark

snd_soc_dai_set_tristate() has added in v2.6.27 (2008/Oct) by this commit

	8c6529dbf881303920a415c2d14a500218661949
	("ALSA: asoc: core - add Digital Audio Interface (DAI) control functions.")

But it has never been used during this 16 years.
Let's remove it.

Kuninori Morimoto (17):
  ASoC: pxa: remove snd_soc_dai_set_tristate()
  ASoC: codecs: cs35l32: remove snd_soc_dai_set_tristate()
  ASoC: codecs: cs35l33: remove snd_soc_dai_set_tristate()
  ASoC: codecs: cs35l34: remove snd_soc_dai_set_tristate()
  ASoC: codecs: cs42l73: remove snd_soc_dai_set_tristate()
  ASoC: codecs: cs53l30: remove snd_soc_dai_set_tristate()
  ASoC: codecs: wm8961: remove snd_soc_dai_set_tristate()
  ASoC: codecs: wm8994: remove snd_soc_dai_set_tristate()
  ASoC: codecs: wm8995: remove snd_soc_dai_set_tristate()
  ASoC: codecs: wm9713: remove snd_soc_dai_set_tristate()
  ASoC: codecs: adau1372: remove snd_soc_dai_set_tristate()
  ASoC: codecs: adau1977: remove snd_soc_dai_set_tristate()
  ASoC: codecs: adau7118: remove snd_soc_dai_set_tristate()
  ASoC: codecs: madera: remove snd_soc_dai_set_tristate()
  ASoC: codecs: arizona: remove snd_soc_dai_set_tristate()
  ASoC: codecs: twl4030: remove snd_soc_dai_set_tristate()
  ASoC: remove snd_soc_dai_set_tristate()

 include/sound/soc-dai.h     |  3 ---
 sound/soc/codecs/adau1372.c | 14 --------------
 sound/soc/codecs/adau1977.c | 15 ---------------
 sound/soc/codecs/adau7118.c | 19 -------------------
 sound/soc/codecs/arizona.c  | 17 -----------------
 sound/soc/codecs/cs35l32.c  |  9 ---------
 sound/soc/codecs/cs35l33.c  | 21 ---------------------
 sound/soc/codecs/cs35l34.c  | 15 ---------------
 sound/soc/codecs/cs42l73.c  | 10 ----------
 sound/soc/codecs/cs53l30.c  | 10 ----------
 sound/soc/codecs/madera.c   | 22 ----------------------
 sound/soc/codecs/twl4030.c  | 28 ----------------------------
 sound/soc/codecs/wm8961.c   | 14 --------------
 sound/soc/codecs/wm8994.c   | 28 ----------------------------
 sound/soc/codecs/wm8995.c   | 33 ---------------------------------
 sound/soc/codecs/wm9713.c   | 17 -----------------
 sound/soc/pxa/pxa-ssp.c     | 21 ---------------------
 sound/soc/soc-dai.c         | 19 -------------------
 18 files changed, 315 deletions(-)

Comments

Charles Keepax Dec. 9, 2024, 9:52 a.m. UTC | #1
On Mon, Dec 09, 2024 at 07:38:52AM +0000, Kuninori Morimoto wrote:
> Hi Mark
> 
> snd_soc_dai_set_tristate() has added in v2.6.27 (2008/Oct) by this commit
> 
> 	8c6529dbf881303920a415c2d14a500218661949
> 	("ALSA: asoc: core - add Digital Audio Interface (DAI) control functions.")
> 
> But it has never been used during this 16 years.
> Let's remove it.

We have definitely used it in out of tree code for some phones in
our times. Whilst I don't feel super strongly, I would say it seems
like fairly reasonable functionality and it does no harm, so I
would vote to keep it.

Thanks,
Charles
Richard Fitzgerald Dec. 9, 2024, 10:26 a.m. UTC | #2
On 09/12/2024 9:52 am, Charles Keepax wrote:
> On Mon, Dec 09, 2024 at 07:38:52AM +0000, Kuninori Morimoto wrote:
>> Hi Mark
>>
>> snd_soc_dai_set_tristate() has added in v2.6.27 (2008/Oct) by this commit
>>
>> 	8c6529dbf881303920a415c2d14a500218661949
>> 	("ALSA: asoc: core - add Digital Audio Interface (DAI) control functions.")
>>
>> But it has never been used during this 16 years.
>> Let's remove it.
> 
> We have definitely used it in out of tree code for some phones in
> our times. Whilst I don't feel super strongly, I would say it seems
> like fairly reasonable functionality and it does no harm, so I
> would vote to keep it.
> 
> Thanks,
> Charles

Yes, I'd forgotten that this function even existed. Now I've been
reminded of it, it's a useful function because our I2S/TDM DAIs have
configurable tri-stating. If we remove the function the alternative
of using a pinmux driver is a lot of overhead.

But the kerneldoc description is not very good:

"Tristates the DAI so that others can use it."

That makes it sound like it disables the DAI. Should it say "Set the DAI
TX pin to be hi-impedance in TDM slots where it is not transmitting"?
Kuninori Morimoto Dec. 10, 2024, 12:06 a.m. UTC | #3
Hi Charles

> > snd_soc_dai_set_tristate() has added in v2.6.27 (2008/Oct) by this commit
> > 
> > 	8c6529dbf881303920a415c2d14a500218661949
> > 	("ALSA: asoc: core - add Digital Audio Interface (DAI) control functions.")
> > 
> > But it has never been used during this 16 years.
> > Let's remove it.
> 
> We have definitely used it in out of tree code for some phones in
> our times. Whilst I don't feel super strongly, I would say it seems
> like fairly reasonable functionality and it does no harm, so I
> would vote to keep it.

Ah, out of tree...
OK, I have no strong opinion if there is a user. let's keep it


Thank you for your help !!

Best regards
---
Kuninori Morimoto