Re: [PATCH 03/14] ASoC: amd: acp: Refactor dmic-codec platform device creation

[potturu venkata prasad <venkataprasad.potturu@amd.com>]

On 3/10/25 17:00, Amadeusz Sławiński wrote:
> On 3/10/2025 11:45 AM, Venkata Prasad Potturu wrote:
>> Refactor dmic-codec platform driver creation using helper function.
>>
>> Signed-off-by: Venkata Prasad Potturu <venkataprasad.potturu@amd.com>
>> ---
>>   sound/soc/amd/acp/acp-pci.c | 49 +++++++++++++++++++++++++------------
>>   sound/soc/amd/acp/amd.h     |  1 +
>>   2 files changed, 35 insertions(+), 15 deletions(-)
>>
>
> ...
>
>>       addr = pci_resource_start(pci, 0);
>>       chip->base = devm_ioremap(&pci->dev, addr, 
>> pci_resource_len(pci, 0));
>>       if (!chip->base) {
>>           ret = -ENOMEM;
>> -        goto unregister_dmic_dev;
>> +        goto release_regions;
>>       }
>>         chip->acp_hw_ops_init(chip);
>>       ret = acp_hw_init(chip);
>>       if (ret)
>> -        goto unregister_dmic_dev;
>> +        goto release_regions;
>
> ...
>
>> @@ -168,8 +187,8 @@ static int acp_pci_probe(struct pci_dev *pci, 
>> const struct pci_device_id *pci_id
>>       pm_runtime_allow(&pci->dev);
>>       return ret;
>>   -unregister_dmic_dev:
>> -    platform_device_unregister(dmic_dev);
>> +de_init:
>> +    chip->acp_hw_ops->acp_deinit(chip);
>
> On init you call acp_hw_init(), but here you call ->acp_deinit() 
> directly instead of acp_hw_deinit()?
Sorry, over looked will fix it in v2 patch series.
>
>>   release_regions:
>>       pci_release_regions(pci);
>>   disable_pci:
>
>