Dne 19. 04. 21 v 9:17 Yang, Libin napsal(a): > Hi Jaroslav, > >> -----Original Message----- >> From: Jaroslav Kysela <perex@xxxxxxxx> >> Sent: 2021年4月19日 14:57 >> To: Yang, Libin <libin.yang@xxxxxxxxx>; Pierre-Louis Bossart <pierre- >> louis.bossart@xxxxxxxxxxxxxxx>; shumingf@xxxxxxxxxxx; broonie@xxxxxxxxxx; >> lgirdwood@xxxxxxxxx >> Cc: oder_chiou@xxxxxxxxxxx; jack.yu@xxxxxxxxxxx; alsa-devel@alsa- >> project.org; lars@xxxxxxxxxx; derek.fang@xxxxxxxxxxx; flove@xxxxxxxxxxx >> Subject: Re: [PATCH] ASoC: rt711-sdca: remove capture switch controls >> >> Dne 19. 04. 21 v 8:50 Yang, Libin napsal(a): >>> Hi Jaroslav, >>> >>>> -----Original Message----- >>>> From: Jaroslav Kysela <perex@xxxxxxxx> >>>> Sent: 2021年4月19日 14:32 >>>> To: Yang, Libin <libin.yang@xxxxxxxxx>; Pierre-Louis Bossart <pierre- >>>> louis.bossart@xxxxxxxxxxxxxxx>; shumingf@xxxxxxxxxxx; >>>> broonie@xxxxxxxxxx; lgirdwood@xxxxxxxxx >>>> Cc: oder_chiou@xxxxxxxxxxx; jack.yu@xxxxxxxxxxx; alsa-devel@alsa- >>>> project.org; lars@xxxxxxxxxx; derek.fang@xxxxxxxxxxx; >>>> flove@xxxxxxxxxxx >>>> Subject: Re: [PATCH] ASoC: rt711-sdca: remove capture switch controls >>>> >>>> Dne 19. 04. 21 v 7:14 Yang, Libin napsal(a): >>>>> Hi Pierre, >>>>> >>>>> >>>>>> -----Original Message----- >>>>>> From: Pierre-Louis Bossart <pierre-louis.bossart@xxxxxxxxxxxxxxx> >>>>>> Sent: 2021年4月15日 19:26 >>>>>> To: shumingf@xxxxxxxxxxx; broonie@xxxxxxxxxx; >> lgirdwood@xxxxxxxxx >>>>>> Cc: oder_chiou@xxxxxxxxxxx; jack.yu@xxxxxxxxxxx; alsa-devel@alsa- >>>>>> project.org; lars@xxxxxxxxxx; Yang, Libin <libin.yang@xxxxxxxxx>; >>>>>> derek.fang@xxxxxxxxxxx; flove@xxxxxxxxxxx >>>>>> Subject: Re: [PATCH] ASoC: rt711-sdca: remove capture switch >>>>>> controls >>>>>> >>>>>> >>>>>> >>>>>> On 4/15/21 4:16 AM, shumingf@xxxxxxxxxxx wrote: >>>>>>> From: Shuming Fan <shumingf@xxxxxxxxxxx> >>>>>>> >>>>>>> The settings of the switch control already set by DAPM event. >>>>>>> These switch controls might the user confused why it can't disable >>>>>>> the >>>>>> capture. >>>>>> >>>>>> Sorry, not following. This control is used in the alsa-ucm-conf >>>>>> that was just merged >>>>>> >>>>>> https://github.com/alsa-project/alsa-ucm- >>>>>> conf/commit/197025656ec456331d1a34357b113913ec3b187f#diff- >>>>>> >>>> >> 0e1c627ea89ee148fdb41aa6b3ba7851ba9c20eb43c1b87b9e0ce92164273dc3R >>>>>> 7 >>>>>> >>>>>> Is this saying this commit needs to be fixed with a follow-up PR? >>>>>> Libin, are you working on this? >>>>> >>>>> Yes, I will submit a patch for UCM after this patch is merged. >>>>> >>>>> The background is Jaroslav requires using codec HW kcontrol for >>>>> capture volume/switch. Shuming and I worked on it. At first we >>>>> wanted to use FU0F to replace PGA kcontrol. But Shuming found FU0F >>>>> is used in >>>> DAPM. >>>>> So it is not proper for the capture volume/switch. And Shuming will >>>>> remove the FU0F kcontrol. >>>> >>>> Is switch control name aligned with the volume control name after >>>> this change? >>> >>> This patch is removing "FU0F Capture Switch". Before this patch, there >>> is "FU0F Capture Switch" and "FU0F Capture Volume". After this patch >>> is applied, "FU0F Capture Switch" is removed. So the UCM >>> https://github.com/alsa-project/alsa-ucm- >> conf/blob/57ead84278f641d411e >>> 3ccbb5c8a4b64141904ba/ucm2/codecs/rt711-sdca/init.conf#L7 >>> of "FU0F Capture Switch" setting needs to be removed. >>> >>>> I mean "A Capture Switch" + "A Capture Volume" not "A Capture Switch" >>>> + "B Capture Volume". >>> >>> We couldn't find a proper codec kcontrol for RT711 capture. So we have >>> to continue using PGA kcontrol. >> >> It's really confusing then. What does "FU0F Capture Volume" ? It's really >> difficult to judge something when I don't know the codec diagram. > > "FU0F Capture Volume" will still control the volume of rt711 capture. > But "FU0F Capture Switch" will be removed. Because DAPM is using > this the same node. These is a conflict when we do below: > 1. mute "FU0F Capture Switch". And the kcontrol will mute the > Corresponding node in codec. > 2. use arecord to capture from the rt711-sdca headset > 3. DAPM will auto unmute the node which "FU0F Capture Switch" has > muted. > 4. So we will capture the sound, even "FU0F Capture Switch" shows the > state is "muted" (this is wrong) > > So Shuming decided to remove the "FU0F Capture Switch" kcontrol. I see. In this case, the auto-route settings should differ from the mixer settings. So the mute flag should be logical _OR_ from both DAPM and the mixer settings. And because the codec is able to do the hw mute, why to prevent the export of this feature? So I propose do do (pseudo code): struct rt711_sdca_priv { bool fu0f_dapm_mute; bool fu0f_mixer_mute; }; /* called from both dapm event and kontrol put callback (on change) */ /* the dapm event and put callback will modify only rt711_sdca_priv fields */ static void set_f0f_mute(rt711_priv) { int mute = rt711_priv->fu0f_dapm_mute || rt711_priv->fu0f_mixer_mute; set_fu0f_mute_register(mute); } With this implementation, all is consistent to the user space. >> Anyway, the switch and volume for the given I/O should have identical name >> and they should differ only in the suffix describing the stream and >> functionality. > > We won't touch "CaptureSwitch" and "CaptureVolume" for rt711-sdca. Yes, but the hw controls should be used instead DSP controls, if they are available. Jaroslav -- Jaroslav Kysela <perex@xxxxxxxx> Linux Sound Maintainer; ALSA Project; Red Hat, Inc.
