snd-rn-pci-acp3x - auto dmic detection unreliable

snd-rn-pci-acp3x - auto dmic detection unreliable

[Jaroslav Kysela]

Hi,

	it seems that the ACPI auto-detection of the presence the digital microphone
connected to AMD Renoir audio bridge is unreliable. It may be that hardware
vendors do not provide the correct ACPI tables.
	Could you clarify the usage of _WOV integer value? It seems that this value
is set on those plaforms which do not have the digital microphone (the
microphone is connected to the HDA codec):

Lenovo IdeaPad 5 (15ARE05)
Lenovo Thinkpad X395

	In any case, we should probably improve the auto-detection using DMI or other
platform info.

					Jaroslav

BugLink: https://bugzilla.redhat.com/show_bug.cgi?id=1892115

-- 
Jaroslav Kysela <perex@perex.cz>
Linux Sound Maintainer; ALSA Project; Red Hat, Inc.

RE: snd-rn-pci-acp3x - auto dmic detection unreliable

[Mukunda, Vijendar]

[AMD Public Use]



-----Original Message-----
From: Jaroslav Kysela <perex@perex.cz> 
Sent: Wednesday, November 18, 2020 2:37 PM
To: Mukunda, Vijendar <Vijendar.Mukunda@amd.com>
Cc: ALSA development <alsa-devel@alsa-project.org>
Subject: snd-rn-pci-acp3x - auto dmic detection unreliable

Hi,

	it seems that the ACPI auto-detection of the presence the digital microphone connected to AMD Renoir audio bridge is unreliable. It may be that hardware vendors do not provide the correct ACPI tables.
	Could you clarify the usage of _WOV integer value? It seems that this value is set on those plaforms which do not have the digital microphone (the microphone is connected to the HDA codec):

_WOV integer value should be set only when DMIC is connected to APU.
When DMIC is not available, _WOV integer value should be set to zero by BIOS.

Lenovo IdeaPad 5 (15ARE05)
Lenovo Thinkpad X395

	In any case, we should probably improve the auto-detection using DMI or other platform info.

					Jaroslav

BugLink: https://nam11.safelinks.protection.outlook.com/?url=https%3A%2F%2Fbugzilla.redhat.com%2Fshow_bug.cgi%3Fid%3D1892115&amp;data=04%7C01%7CVijendar.Mukunda%40amd.com%7C4d901531b6454bebc58208d88ba15a4b%7C3dd8961fe4884e608e11a82d994e183d%7C0%7C0%7C637412872412887848%7CUnknown%7CTWFpbGZsb3d8eyJWIjoiMC4wLjAwMDAiLCJQIjoiV2luMzIiLCJBTiI6Ik1haWwiLCJXVCI6Mn0%3D%7C1000&amp;sdata=S7769uYGT3db%2BaztzZyobmVN2leseXPqjR6pktFUaR0%3D&amp;reserved=0

--
Jaroslav Kysela <perex@perex.cz>
Linux Sound Maintainer; ALSA Project; Red Hat, Inc.

Re: snd-rn-pci-acp3x - auto dmic detection unreliable

[Jaroslav Kysela]

Dne 18. 11. 20 v 10:25 Mukunda, Vijendar napsal(a):
> [AMD Public Use]

AMD: You just replicated my original question without any answer (2 weeks+).
Do you have something to share or do you not support your code?

				Jaroslav

> -----Original Message-----
> From: Jaroslav Kysela <perex@perex.cz> 
> Sent: Wednesday, November 18, 2020 2:37 PM
> To: Mukunda, Vijendar <Vijendar.Mukunda@amd.com>
> Cc: ALSA development <alsa-devel@alsa-project.org>
> Subject: snd-rn-pci-acp3x - auto dmic detection unreliable
> 
> Hi,
> 
> 	it seems that the ACPI auto-detection of the presence the digital microphone connected to AMD Renoir audio bridge is unreliable. It may be that hardware vendors do not provide the correct ACPI tables.
> 	Could you clarify the usage of _WOV integer value? It seems that this value is set on those plaforms which do not have the digital microphone (the microphone is connected to the HDA codec):
> 
> _WOV integer value should be set only when DMIC is connected to APU.
> When DMIC is not available, _WOV integer value should be set to zero by BIOS.
> 
> Lenovo IdeaPad 5 (15ARE05)
> Lenovo Thinkpad X395
> 
> 	In any case, we should probably improve the auto-detection using DMI or other platform info.
> 
> 					Jaroslav
> 
> BugLink: https://nam11.safelinks.protection.outlook.com/?url=https%3A%2F%2Fbugzilla.redhat.com%2Fshow_bug.cgi%3Fid%3D1892115&amp;data=04%7C01%7CVijendar.Mukunda%40amd.com%7C4d901531b6454bebc58208d88ba15a4b%7C3dd8961fe4884e608e11a82d994e183d%7C0%7C0%7C637412872412887848%7CUnknown%7CTWFpbGZsb3d8eyJWIjoiMC4wLjAwMDAiLCJQIjoiV2luMzIiLCJBTiI6Ik1haWwiLCJXVCI6Mn0%3D%7C1000&amp;sdata=S7769uYGT3db%2BaztzZyobmVN2leseXPqjR6pktFUaR0%3D&amp;reserved=0
> 
> --
> Jaroslav Kysela <perex@perex.cz>
> Linux Sound Maintainer; ALSA Project; Red Hat, Inc.
> 


-- 
Jaroslav Kysela <perex@perex.cz>
Linux Sound Maintainer; ALSA Project; Red Hat, Inc.

Re: snd-rn-pci-acp3x - auto dmic detection unreliable

[Jaroslav Kysela]

Dne 07. 12. 20 v 10:51 Mukunda,Vijendar napsal(a):
> 
> Corrected typo error.
> 
> On 07/12/20 3:19 pm, Mukunda,Vijendar wrote:
>>
>>
>> On 07/12/20 2:22 pm, Jaroslav Kysela wrote:
>>> Dne 18. 11. 20 v 10:25 Mukunda, Vijendar napsal(a):
>>>> [AMD Public Use]
>>>
>>> AMD: You just replicated my original question without any answer (2 
>>> weeks+).
>>> Do you have something to share or do you not support your code?
>>
>> We are already provided reply on 18th Nov.
>>
>> _WOV integer value should be set only when DMIC is connected to APU.
>> When DMIC is not connected to APU,  _WOV integer value should be set to 
>> zero by BIOS.
>>
>>   DMIC doesn't had auto presence detect logic on AMD platforms and DMIC 
>> connection to AMD platform is simple wiring.
>> Whether DMIC connected to APU or HDAUDIO codec can be decided by ODMs.
>>
> ODM BIOS has to update the _WOV method value based on DMIC connected to APU
>> or not.

Thank you. Please, format correctly your e-mail messages, do not write your
answers inside the old text without '>' markers:

https://mailman.alsa-project.org/pipermail/alsa-devel/2020-November/177218.html

I just overlooked your lines inside my forwarded message.

So the problem is that the hardware vendors do not do the right job in this case.

			Thank you again,
					Jaroslav

-- 
Jaroslav Kysela <perex@perex.cz>
Linux Sound Maintainer; ALSA Project; Red Hat, Inc.

Re: snd-rn-pci-acp3x - auto dmic detection unreliable

[Mukunda,Vijendar]

On 07/12/20 2:22 pm, Jaroslav Kysela wrote:
> Dne 18. 11. 20 v 10:25 Mukunda, Vijendar napsal(a):
>> [AMD Public Use]
> 
> AMD: You just replicated my original question without any answer (2 weeks+).
> Do you have something to share or do you not support your code?

We are already provided reply on 18th Nov.

_WOV integer value should be set only when DMIC is connected to APU.
When DMIC is not connected to APU,  _WOV integer value should be set to 
zero by BIOS.

  DMIC doesn't had auto presence detect logic on AMD platforms and DMIC 
connection to AMD platform is simple wiring.
Whether DMIC connected to APU or HDAUDIO codec can be decided by ODMs.

ODM BIOS has update the _WOV method value based on DMIC connected to APU 
or not.

-
Vijendar



> 
> 				Jaroslav
> 
>> -----Original Message-----
>> From: Jaroslav Kysela <perex@perex.cz>
>> Sent: Wednesday, November 18, 2020 2:37 PM
>> To: Mukunda, Vijendar <Vijendar.Mukunda@amd.com>
>> Cc: ALSA development <alsa-devel@alsa-project.org>
>> Subject: snd-rn-pci-acp3x - auto dmic detection unreliable
>>
>> Hi,
>>
>> 	it seems that the ACPI auto-detection of the presence the digital microphone connected to AMD Renoir audio bridge is unreliable. It may be that hardware vendors do not provide the correct ACPI tables.
>> 	Could you clarify the usage of _WOV integer value? It seems that this value is set on those plaforms which do not have the digital microphone (the microphone is connected to the HDA codec):
>>
>> _WOV integer value should be set only when DMIC is connected to APU.
>> When DMIC is not available, _WOV integer value should be set to zero by BIOS.

We are using
>>
>> Lenovo IdeaPad 5 (15ARE05)
>> Lenovo Thinkpad X395
>>
>> 	In any case, we should probably improve the auto-detection using DMI or other platform info.
>>
>> 					Jaroslav
>>
>> BugLink: https://nam11.safelinks.protection.outlook.com/?url=https%3A%2F%2Fbugzilla.redhat.com%2Fshow_bug.cgi%3Fid%3D1892115&amp;data=04%7C01%7CVijendar.Mukunda%40amd.com%7C99b9a10b76674aa6a28a08d89a8d75fc%7C3dd8961fe4884e608e11a82d994e183d%7C0%7C0%7C637429279661047562%7CUnknown%7CTWFpbGZsb3d8eyJWIjoiMC4wLjAwMDAiLCJQIjoiV2luMzIiLCJBTiI6Ik1haWwiLCJXVCI6Mn0%3D%7C1000&amp;sdata=CpZLrKf4pVdN6MmnmWUnXCLevFjySGuFW%2BJeLoYW%2BjQ%3D&amp;reserved=0
>>
>> --
>> Jaroslav Kysela <perex@perex.cz>
>> Linux Sound Maintainer; ALSA Project; Red Hat, Inc.
>>
> 
> 

Re: snd-rn-pci-acp3x - auto dmic detection unreliable

[Mukunda,Vijendar]

Corrected typo error.

On 07/12/20 3:19 pm, Mukunda,Vijendar wrote:
> 
> 
> On 07/12/20 2:22 pm, Jaroslav Kysela wrote:
>> Dne 18. 11. 20 v 10:25 Mukunda, Vijendar napsal(a):
>>> [AMD Public Use]
>>
>> AMD: You just replicated my original question without any answer (2 
>> weeks+).
>> Do you have something to share or do you not support your code?
> 
> We are already provided reply on 18th Nov.
> 
> _WOV integer value should be set only when DMIC is connected to APU.
> When DMIC is not connected to APU,  _WOV integer value should be set to 
> zero by BIOS.
> 
>   DMIC doesn't had auto presence detect logic on AMD platforms and DMIC 
> connection to AMD platform is simple wiring.
> Whether DMIC connected to APU or HDAUDIO codec can be decided by ODMs.
> 
ODM BIOS has to update the _WOV method value based on DMIC connected to APU
> or not.
> 
> -
> Vijendar
> 
> 
> 
>>
>>                 Jaroslav
>>
>>> -----Original Message-----
>>> From: Jaroslav Kysela <perex@perex.cz>
>>> Sent: Wednesday, November 18, 2020 2:37 PM
>>> To: Mukunda, Vijendar <Vijendar.Mukunda@amd.com>
>>> Cc: ALSA development <alsa-devel@alsa-project.org>
>>> Subject: snd-rn-pci-acp3x - auto dmic detection unreliable
>>>
>>> Hi,
>>>
>>>     it seems that the ACPI auto-detection of the presence the digital 
>>> microphone connected to AMD Renoir audio bridge is unreliable. It may 
>>> be that hardware vendors do not provide the correct ACPI tables.
>>>     Could you clarify the usage of _WOV integer value? It seems that 
>>> this value is set on those plaforms which do not have the digital 
>>> microphone (the microphone is connected to the HDA codec):
>>>
>>> _WOV integer value should be set only when DMIC is connected to APU.
>>> When DMIC is not available, _WOV integer value should be set to zero 
>>> by BIOS.
> 
> We are using
>>>
>>> Lenovo IdeaPad 5 (15ARE05)
>>> Lenovo Thinkpad X395
>>>
>>>     In any case, we should probably improve the auto-detection using 
>>> DMI or other platform info.
>>>
>>>                     Jaroslav
>>>
>>> BugLink: 
>>> https://nam11.safelinks.protection.outlook.com/?url=https%3A%2F%2Fbugzilla.redhat.com%2Fshow_bug.cgi%3Fid%3D1892115&amp;data=04%7C01%7CVijendar.Mukunda%40amd.com%7C99b9a10b76674aa6a28a08d89a8d75fc%7C3dd8961fe4884e608e11a82d994e183d%7C0%7C0%7C637429279661047562%7CUnknown%7CTWFpbGZsb3d8eyJWIjoiMC4wLjAwMDAiLCJQIjoiV2luMzIiLCJBTiI6Ik1haWwiLCJXVCI6Mn0%3D%7C1000&amp;sdata=CpZLrKf4pVdN6MmnmWUnXCLevFjySGuFW%2BJeLoYW%2BjQ%3D&amp;reserved=0 
>>>
>>>
>>> -- 
>>> Jaroslav Kysela <perex@perex.cz>
>>> Linux Sound Maintainer; ALSA Project; Red Hat, Inc.
>>>
>>
>>

Re: snd-rn-pci-acp3x - auto dmic detection unreliable

[Mukunda,Vijendar]

On 07/12/20 3:15 pm, Jaroslav Kysela wrote:
> Dne 07. 12. 20 v 10:51 Mukunda,Vijendar napsal(a):
>>
>> Corrected typo error.
>>
>> On 07/12/20 3:19 pm, Mukunda,Vijendar wrote:
>>>
>>>
>>> On 07/12/20 2:22 pm, Jaroslav Kysela wrote:
>>>> Dne 18. 11. 20 v 10:25 Mukunda, Vijendar napsal(a):
>>>>> [AMD Public Use]
>>>>
>>>> AMD: You just replicated my original question without any answer (2
>>>> weeks+).
>>>> Do you have something to share or do you not support your code?
>>>
>>> We are already provided reply on 18th Nov.
>>>
>>> _WOV integer value should be set only when DMIC is connected to APU.
>>> When DMIC is not connected to APU,  _WOV integer value should be set to
>>> zero by BIOS.
>>>
>>>    DMIC doesn't had auto presence detect logic on AMD platforms and DMIC
>>> connection to AMD platform is simple wiring.
>>> Whether DMIC connected to APU or HDAUDIO codec can be decided by ODMs.
>>>
>> ODM BIOS has to update the _WOV method value based on DMIC connected to APU
>>> or not.
> 
> Thank you. Please, format correctly your e-mail messages, do not write your
> answers inside the old text without '>' markers:

Sorry it's my bad. Due to limited access to my linux machine, Last time 
I have replied from Outlook server instead of thunder bird mail client 
which resulted format related issues.

-
Vijendar


> 
> https://nam11.safelinks.protection.outlook.com/?url=https%3A%2F%2Fmailman.alsa-project.org%2Fpipermail%2Falsa-devel%2F2020-November%2F177218.html&data=04%7C01%7Cvijendar.mukunda%40amd.com%7C9aec4eab94f2443adc1e08d89a94cfd6%7C3dd8961fe4884e608e11a82d994e183d%7C0%7C0%7C637429311243402463%7CUnknown%7CTWFpbGZsb3d8eyJWIjoiMC4wLjAwMDAiLCJQIjoiV2luMzIiLCJBTiI6Ik1haWwiLCJXVCI6Mn0%3D%7C1000&sdata=wn7KBu8KbQOWaR9PEnyDwbLR%2B%2B86Z2xFRAnyUy%2FN3YQ%3D&reserved=0
> 
> I just overlooked your lines inside my forwarded message.
> 
> So the problem is that the hardware vendors do not do the right job in this case.
> 
> 			Thank you again,
> 					Jaroslav
>