Re: [PATCH 5/5] ASoC: amd: acp7x: add system and runtime PM ops

On 5/21/26 06:41, Mukunda,Vijendar wrote:
> On 08/05/26 01:03, Mario Limonciello wrote:
>>
>>
>> On 5/7/26 13:11, Vijendar Mukunda wrote:
>>> Add ACP7.x PM callbacks and hook up system sleep and runtime PM ops for
>>> the PCI driver.
>>>
>>> Signed-off-by: Vijendar Mukunda <[email protected]>
>>> ---
>>>    sound/soc/amd/acp7x/acp7x-common.c | 43 ++++++++++++++++++++++++++++++
>>>    sound/soc/amd/acp7x/acp7x.h        |  3 +++
>>>    sound/soc/amd/acp7x/pci-acp7x.c    | 31 +++++++++++++++++++++
>>>    3 files changed, 77 insertions(+)
>>>
>>> diff --git a/sound/soc/amd/acp7x/acp7x-common.c
>>> b/sound/soc/amd/acp7x/acp7x-common.c
>>> index 68f9b47766c4..df94864a1693 100644
>>> --- a/sound/soc/amd/acp7x/acp7x-common.c
>>> +++ b/sound/soc/amd/acp7x/acp7x-common.c
>>> @@ -75,8 +75,51 @@ static int acp7x_deinit(void __iomem *acp_base, struct
>>> device *dev)
>>>        return 0;
>>>    }
>>>    +static int __maybe_unused snd_acp7x_suspend(struct device *dev)
>>> +{
>>> +    struct acp7x_dev_data *adata;
>>> +    int ret;
>>> +
>>> +    adata = dev_get_drvdata(dev);
>>> +    ret = acp_hw_deinit(adata, dev);
>>> +    if (ret)
>>> +        dev_err(dev, "ACP de-init failed\n");
>>> +    return ret;
>>> +}
>>> +
>>> +static int __maybe_unused snd_acp7x_runtime_resume(struct device *dev)
>>> +{
>>> +    struct acp7x_dev_data *adata;
>>> +    int ret;
>>> +
>>> +    adata = dev_get_drvdata(dev);
>>> +    ret = acp_hw_init(adata, dev);
>>> +    if (ret) {
>>> +        dev_err(dev, "ACP init failed\n");
>>> +        return ret;
>>> +    }
>>> +    return 0;
>>> +}
>>> +
>>> +static int __maybe_unused snd_acp7x_resume(struct device *dev)
>>> +{
>>> +    struct acp7x_dev_data *adata;
>>> +    int ret;
>>> +
>>> +    adata = dev_get_drvdata(dev);
>>> +    ret = acp_hw_init(adata, dev);
>>> +    if (ret)
>>> +        dev_err(dev, "ACP init failed\n");
>>> +
>>> +    return ret;
>>> +}
>>
>> Both of snd_acp7x_runtime_resume and snd_acp7x_resume look identical. Will
>> there be later code (when machine driver and IO enablement come) that makes
>> them different?
> 
> Yes, when IO enablement code is updated, both the functions are going to be
> different.
> We need to keep it separately.
> 

Thanks, no concerns with the series then.

Reviewed-by: Mario Limonciello (AMD) <[email protected]>