Re: [PATCH] ASoC: Intel: sof_sdw: initialize ret in asoc_sdw_rt_amp_spk_rtd_init()

[Pierre-Louis Bossart <pierre-louis.bossart@xxxxxxxxx>]

On 2/11/25 07:13, Mark Brown wrote:
> On Mon, Feb 10, 2025 at 11:08:27PM -0500, Ethan Carter Edwards wrote:
>> There is a possibility for an uninitialized *ret* variable to be
>> returned in some code paths.
>>
>> Setting to 0 prevents a random value from being returned.
> 
> That'll shut up the warning but is the warning trying to tell us that
> there's a logic bug somewhere in the function and we're for example
> forgetting to look at a return value in some path in the function?

The problematic code is this:

for_each_rtd_codec_dais(rtd, i, codec_dai) {
	if (strstr(codec_dai->component->name_prefix, "-1"))
		ret = snd_soc_dapm_add_routes(&card->dapm, rt_amp_map, 2);
	else if (strstr(codec_dai->component->name_prefix, "-2"))
		ret = snd_soc_dapm_add_routes(&card->dapm, rt_amp_map + 2, 2);
}

return ret;

I am not sure if it's possible that either the for_each does nothing or that the two branches are skipped, but certainly initializing the 'ret' value makes sense to me.

Bard, Shuming, what do you think?