Message-ID: <ac7202f4-e5c8-477b-b805-685f573d179a@linux.intel.com>
Date: Tue, 4 Nov 2025 08:14:13 +0200
From: Péter Ujfalusi <peter.ujfalusi@...ux.intel.com>
To: Bjorn Helgaas <helgaas@...nel.org>
Cc: lgirdwood@...il.com, broonie@...nel.org, linux-sound@...r.kernel.org,
 kai.vehmanen@...ux.intel.com, ranjani.sridharan@...ux.intel.com,
 yung-chuan.liao@...ux.intel.com, pierre-louis.bossart@...ux.dev,
 bhelgaas@...gle.com, linux-pci@...r.kernel.org,
 linux-kernel@...r.kernel.org, kw@...ux.com
Subject: Re: [PATCH 1/7] PCI: Add Intel Nova Lake S audio Device ID



On 03/11/2025 19:33, Bjorn Helgaas wrote:
> On Mon, Nov 03, 2025 at 06:27:16PM +0200, Péter Ujfalusi wrote:
>> On 03/11/2025 18:02, Bjorn Helgaas wrote:
>>> On Mon, Nov 03, 2025 at 02:43:57PM +0200, Péter Ujfalusi wrote:
>>>> On 02/10/2025 11:42, Peter Ujfalusi wrote:
>>>>> Add Nova Lake S (NVL-S) audio Device ID
>>>>
>>>> Can you check this patch so Takashi-san can pick the series up?
>>>
>>> We have a long history of adding these Intel audio device IDs that are
>>> only used once, which is not our usual practice per the comment at the
>>> top of the file:
>>>
>>>  *      Do not add new entries to this file unless the definitions
>>>  *      are shared between multiple drivers.
>>>
>>> Generally speaking, if an ID is used by only a single driver, we
>>> either use the plain hex ID or add the #define to the driver that uses
>>> it.
>>
>> In this case the ID is used by two different driver stack, the
>> legacy HDA and SOF.
> 
> Sigh.  I looked through the patch series, searching for
> PCI_DEVICE_ID_INTEL_HDA_NVL_S, but of course there's only one instance
> of *that*, but two others constructed via PCI_DEVICE_DATA() where only
> "HDA_NVL_S" is mentioned.

I'm not sure if it would be better, but should we move the HDA PCI IDs
to an audio specific header?
Like include/sound/hda_pci_ids.h
It looks to me that mostly if not only these are Intel IDs.

Not in this series, but as a separate one.
> Can you include some hint about that in the commit log so I don't have
> to go through this whole exercise every time?  I want pci_ids.h
> changes to mention the multiple places a new ID is used so I know that
> the "multiple uses" rule has been observed.
> 
> With that:
> 
> Acked-by: Bjorn Helgaas <bhelgaas@...gle.com>

Thank you, I will send v2 with this update.>
>>> Have we been operating under some special exception for the Intel
>>> audio IDs?  I see that I acked some of these additions in the past,
>>> but I don't remember why.
>>
>> The HDA audio entries were moved here by v4 of this series:
>> https://www.spinics.net/lists/alsa-devel/msg161995.html
>>
>> (I cannot find link to v4, only this:
>> https://patchwork.ozlabs.org/project/linux-pci/list/?series=364212)
>>>>> Signed-off-by: Peter Ujfalusi <peter.ujfalusi@...ux.intel.com>
>>>>> Reviewed-by: Kai Vehmanen <kai.vehmanen@...ux.intel.com>
>>>>> Reviewed-by: Ranjani Sridharan <ranjani.sridharan@...ux.intel.com>
>>>>> ---
>>>>>  include/linux/pci_ids.h | 1 +
>>>>>  1 file changed, 1 insertion(+)
>>>>>
>>>>> diff --git a/include/linux/pci_ids.h b/include/linux/pci_ids.h
>>>>> index 92ffc4373f6d..a9a089566b7c 100644
>>>>> --- a/include/linux/pci_ids.h
>>>>> +++ b/include/linux/pci_ids.h
>>>>> @@ -3075,6 +3075,7 @@
>>>>>  #define PCI_DEVICE_ID_INTEL_5100_22	0x65f6
>>>>>  #define PCI_DEVICE_ID_INTEL_IOAT_SCNB	0x65ff
>>>>>  #define PCI_DEVICE_ID_INTEL_HDA_FCL	0x67a8
>>>>> +#define PCI_DEVICE_ID_INTEL_HDA_NVL_S	0x6e50
>>>>>  #define PCI_DEVICE_ID_INTEL_82371SB_0	0x7000
>>>>>  #define PCI_DEVICE_ID_INTEL_82371SB_1	0x7010
>>>>>  #define PCI_DEVICE_ID_INTEL_82371SB_2	0x7020
>>>>
>>>>
>>
>> -- 
>> Péter
>>

-- 
Péter

Powered by blists - more mailing lists