[AMD Official Use Only - Internal Distribution Only] So Actually for rt5682 codec Now in 5.8 there are three flags : SND_SOC_RT5682 SND_SOC_RT5682_I2C SND_SOC_RT5682_SDW But till 5.7.8 we have SND_SOC_RT5682 SND_SOC_RT5682_SDW So in our design we were using SND_SOC_RT5682 which build snd_soc_rt5682.ko Creates the respective codec_dais as defined in that .ko If we use SND_SOC_RT5682_I2C we get snd_soc_rt5682_I2c.ko , it is not creating the expected codec_dai links. As there are three flags defined in codecs, I expect that previous one which we were using(SND_SOC_RT5682) is not a wrong flag and I expect to use SND_SOC_RT5682 as it is still available. Thanks, Vishnu -----Original Message----- From: Mark Brown <broonie@xxxxxxxxxx> Sent: Monday, July 27, 2020 9:40 PM To: Pierre-Louis Bossart <pierre-louis.bossart@xxxxxxxxxxxxxxx> Cc: RAVULAPATI, VISHNU VARDHAN RAO <Vishnuvardhanrao.Ravulapati@xxxxxxx>; moderated list:SOUND - SOC LAYER / DYNAMIC AUDIO POWER MANAGEM... <alsa-devel@xxxxxxxxxxxxxxxx>; Arnd Bergmann <arnd@xxxxxxxx>; open list <linux-kernel@xxxxxxxxxxxxxxx>; YueHaibing <yuehaibing@xxxxxxxxxx>; Takashi Iwai <tiwai@xxxxxxxx>; Enric Balletbo i Serra <enric.balletbo@xxxxxxxxxxxxx>; Liam Girdwood <lgirdwood@xxxxxxxxx>; Mukunda, Vijendar <Vijendar.Mukunda@xxxxxxx>; Deucher, Alexander <Alexander.Deucher@xxxxxxx>; Agrawal, Akshu <Akshu.Agrawal@xxxxxxx> Subject: Re: [PATCH 3/6] ASoC: amd: SND_SOC_RT5682_I2C does not build rt5682 On Mon, Jul 27, 2020 at 10:31:24AM -0500, Pierre-Louis Bossart wrote: > On 7/27/20 9:58 AM, Ravulapati Vishnu vardhan rao wrote: > > changing SND_SOC_RT5682_I2C to SND_SOC_RT5682 because, This flag > > which was previously set as SND_SOC_RT5682 used to build rt5682 > > codec driver but by changing into SND_SOC_RT5682_I2C is preventing > > to build rt5682 codec driver and machine driver fails to probe.So > > Reverting the changes. > The split between I2C and SoundWire means you have to choose the I2C > or SDW mode. Selecting the common part looks very strange. Right, and I can't understand the commit message at all. What's the actual issue here and how could this fix it - in what situation wouldn't you need one of the bus modules?