Re: [PATCH 7/9] ASoC: wm8978: stop resume after register restore errors

[Charles Keepax]

On Sun, Sep 06, 2026 at 11:43:19AM +0800, Pengpeng Hou wrote:
> wm8978_resume() ignores failures from cache replay, bias restoration and
> PLL re-enable. It can therefore continue publishing later resume steps
> after the codec restore transaction has already failed.
> 
> Return the first negative result from those ordered operations. Treat
> the positive "register changed" result from
> snd_soc_component_update_bits() as success, as required by the component
> resume callback contract. The ASoC wrapper reports failures and retains
> best-effort card resume.
> 
> The issue was found by our static-analysis tool and manually reviewed.
> 
> Fixes: 0d34e91596ef ("ASoC: add a WM8978 codec driver")
> Assisted-by: gpt 5
> Signed-off-by: Pengpeng Hou <hppiscas@xxxxxxx>
> ---
>  sound/soc/codecs/wm8978.c | 18 ++++++++++++++----
>  1 file changed, 14 insertions(+), 4 deletions(-)
> 
> diff --git a/sound/soc/codecs/wm8978.c b/sound/soc/codecs/wm8978.c
> index ad8064bbaaac..241d17d81ecb 100644
> --- a/sound/soc/codecs/wm8978.c
> +++ b/sound/soc/codecs/wm8978.c
> @@ -940,15 +940,25 @@ static int wm8978_resume(struct snd_soc_component *component)
>  {
>  	struct wm8978_priv *wm8978 = snd_soc_component_get_drvdata(component);
>  	struct snd_soc_dapm_context *dapm = snd_soc_component_to_dapm(component);
> +	int ret;
>  
>  	/* Sync reg_cache with the hardware */
> -	regcache_sync(wm8978->regmap);
> +	ret = regcache_sync(wm8978->regmap);
> +	if (ret)
> +		return ret;
>  
> -	snd_soc_dapm_force_bias_level(dapm, SND_SOC_BIAS_STANDBY);
> +	ret = snd_soc_dapm_force_bias_level(dapm, SND_SOC_BIAS_STANDBY);
> +	if (ret)
> +		return ret;
>  
> -	if (wm8978->f_pllout)
> +	if (wm8978->f_pllout) {
>  		/* Switch PLL on */
> -		snd_soc_component_update_bits(component, WM8978_POWER_MANAGEMENT_1, 0x20, 0x20);
> +		ret = snd_soc_component_update_bits(component,
> +						    WM8978_POWER_MANAGEMENT_1,
> +						    0x20, 0x20);
> +		if (ret < 0)
> +			return ret;

What is the thinking on not pushing the bias_level back to off
here?

Thanks,
Charles