[PATCH v2 2/7] ASoC: dt-bindings: mediatek,mt8188-mt6359: remove ADDA_BE from link-name
Alexandre Mergnat
amergnat at baylibre.com
Wed May 24 06:28:38 PDT 2023
On 24/05/2023 04:25, Trevor Wu (吳文良) wrote:
> On Tue, 2023-05-23 at 18:26 +0200, Alexandre Mergnat wrote:
>> On 23/05/2023 04:19, Trevor Wu wrote:
>>> ADDA_BE is used to connect to mt6359. For machine mt8188-mt6359,
>>> codec
>>> for ADDA_BE must be mt6359 which are configured on the machine
>>> driver.
>>> Besides, ADDA_BE is divided into two dais, UL_SRC_BE and DL_SRC_BE.
>>> As a result, remove ADDA_BE from items of link-name.
>>>
>>> Signed-off-by: Trevor Wu<trevor.wu at mediatek.com>
>>
>> I don't understand how "DL_SRC_BE" and "UL_SRC_BE" links are done.
>> Why these dais don't replace "ADDA_BE" in this binding ?
>>
>> Regards,
>> Alexandre
>>
>
> Hi Alexandre,
>
> Because the sound card is mt8188-mt6359, the codec for these two links
> must be mt6359. Thus, I specifiy the codec in machine driver directly.
> If the codec is changed, there will be a new sound card and binding
> file. In conclusion, the codec won't be updated via dts, and that's why
> I don't just replace ADDA_BE in this binding.
>
> Do you suggest me add some information in the commit message?
No it's fine, I'm just trying to understand.
When you say "I specifiy the codec in machine driver directly", you
are talking about this change ?
+ } else if (strcmp(dai_link->name, "DL_SRC_BE") == 0 ||
+ strcmp(dai_link->name, "UL_SRC_BE") == 0) {
+ if (!init_mt6359) {
+ dai_link->init = mt8188_mt6359_init;
I'm asking because the equivalent was done here:
- [DAI_LINK_ADDA_BE] = {
- .name = "ADDA_BE",
+ [DAI_LINK_DL_SRC_BE] = {
+ .name = "DL_SRC_BE",
.no_pcm = 1,
.dpcm_playback = 1,
- .dpcm_capture = 1,
- .init = mt8188_mt6359_init,
- SND_SOC_DAILINK_REG(adda),
+ SND_SOC_DAILINK_REG(dl_src),
So I'm wondering why "ADDA_BE" & "DPTX_BE" & "ETDM3_OUT_BE" are in the
enum list of the binding since the codec is already specified in
machine driver too. I probably miss something but I don't know what.
--
Regards,
Alexandre
More information about the linux-arm-kernel
mailing list