[alsa-devel] [PATCH 4/5] ASoC: sam9x5_wm8731: Fix OF node reference counting

[Ladislav Michl]

On Mon, Jan 15, 2018 at 09:46:07AM +0100, Alexandre Belloni wrote:
> On 15/01/2018 at 08:54:30 +0100, Ladislav Michl wrote:
> > of_node_release() is called as the destructor on a dynamically
> > allocated node in of_node_put(), but node pointer is still in use.
> > Fix that by moving of_node_put() into the error path.
> > 
> > Signed-off-by: Ladislav Michl <ladis at linux-mips.org>
> > ---
> >  sound/soc/atmel/sam9x5_wm8731.c | 27 ++++++++++++---------------
> >  1 file changed, 12 insertions(+), 15 deletions(-)
> > 
> > diff --git a/sound/soc/atmel/sam9x5_wm8731.c b/sound/soc/atmel/sam9x5_wm8731.c
> > index 9db08826b32a..bcfb474ee0e8 100644
> > --- a/sound/soc/atmel/sam9x5_wm8731.c
> > +++ b/sound/soc/atmel/sam9x5_wm8731.c
> > @@ -78,7 +78,6 @@ static const struct snd_soc_dapm_widget sam9x5_dapm_widgets[] = {
> >  static int sam9x5_wm8731_driver_probe(struct platform_device *pdev)
> >  {
> >  	struct device_node *np = pdev->dev.of_node;
> > -	struct device_node *codec_np, *cpu_np;
> >  	struct snd_soc_card *card;
> >  	struct snd_soc_dai_link *dai;
> >  	struct sam9x5_drvdata *priv;
> > @@ -122,36 +121,30 @@ static int sam9x5_wm8731_driver_probe(struct platform_device *pdev)
> >  		goto out;
> >  	}
> >  
> > -	codec_np = of_parse_phandle(np, "atmel,audio-codec", 0);
> > -	if (!codec_np) {
> > +	dai->codec_of_node = of_parse_phandle(np, "atmel,audio-codec", 0);
> > +	if (!dai->codec_of_node) {
> >  		dev_err(&pdev->dev, "atmel,audio-codec node missing\n");
> >  		ret = -EINVAL;
> >  		goto out;
> >  	}
> >  
> > -	dai->codec_of_node = codec_np;
> > -
> > -	cpu_np = of_parse_phandle(np, "atmel,ssc-controller", 0);
> > -	if (!cpu_np) {
> > +	dai->cpu_of_node = of_parse_phandle(np, "atmel,ssc-controller", 0);
> > +	if (!dai->cpu_of_node) {
> >  		dev_err(&pdev->dev, "atmel,ssc-controller node missing\n");
> >  		ret = -EINVAL;
> > -		goto out;
> > +		goto out_put_codec;
> >  	}
> > -	dai->cpu_of_node = cpu_np;
> > -	dai->platform_of_node = cpu_np;
> > +	dai->platform_of_node = dai->cpu_of_node;
> >  
> > -	priv->ssc_id = of_alias_get_id(cpu_np, "ssc");
> > +	priv->ssc_id = of_alias_get_id(dai->cpu_of_node, "ssc");
> >  
> >  	ret = atmel_ssc_set_audio(priv->ssc_id);
> >  	if (ret != 0) {
> >  		dev_err(&pdev->dev, "Failed to set SSC %d for audio: %d\n",
> >  			ret, priv->ssc_id);
> > -		goto out;
> > +		goto out_put_cpu;
> >  	}
> >  
> > -	of_node_put(codec_np);
> > -	of_node_put(cpu_np);
> > -
> >  	ret = snd_soc_register_card(card);
> >  	if (ret) {
> >  		dev_err(&pdev->dev, "Platform device allocation failed\n");
> > @@ -164,6 +157,10 @@ static int sam9x5_wm8731_driver_probe(struct platform_device *pdev)
> >  
> >  out_put_audio:
> >  	atmel_ssc_put_audio(priv->ssc_id);
> > +out_put_cpu:
> > +	of_node_put(dai->cpu_of_node);
> 
> Should'nt they be put in sam9x5_wm8731_driver_remove then?

No, sound core should do clean up as it is done in
asoc_simple_card_clean_reference. I didn't read whole related ALSA code (yet).
Perhaps core developers could bring a bit of light here?

> Did you test any of those changes?

No as I do not have hardware :)

> > +out_put_codec:
> > +	of_node_put(dai->codec_of_node);
> >  out:
> >  	return ret;
> >  }
> > -- 
> > 2.15.1
> > 
> 
> -- 
> Alexandre Belloni, Free Electrons
> Embedded Linux and Kernel engineering
> http://free-electrons.com