| Message ID | 20220325154241.1600757-3-broonie@kernel.org |
|---|---|
| State | Accepted |
| Commit | 28103509248b94392e04a8ffcbc47da5e3e31dfc |
| Headers | show |
| Series | ASoC: atmel: Fixes for AT91SAM9G20-EK audio driver | expand |
On 25.03.2022 17:42, Mark Brown wrote: > The error handling in the AT91SAM9G20-EK machine driver probe did not > consistently free the SSC in error paths, sometimes immediately returning > an error rather than doing cleanup. Fix this. > > Signed-off-by: Mark Brown <broonie@kernel.org> Should we have a 'Fixes' tag here? Other than that: Reviewed-by: Codrin Ciubotariu <codrin.ciubotariu@microchip.com> Thanks!
On Mon, Apr 04, 2022 at 12:47:27PM +0000, Codrin.Ciubotariu@microchip.com wrote: > On 25.03.2022 17:42, Mark Brown wrote: > > The error handling in the AT91SAM9G20-EK machine driver probe did not > > consistently free the SSC in error paths, sometimes immediately returning > > an error rather than doing cleanup. Fix this. > > > > Signed-off-by: Mark Brown <broonie@kernel.org> > > Should we have a 'Fixes' tag here? I'm not sure those code paths are ever executing, I really don't think it's worth the effort or noise checking.
diff --git a/sound/soc/atmel/sam9g20_wm8731.c b/sound/soc/atmel/sam9g20_wm8731.c index 0d639a33ad96..d771843011ea 100644 --- a/sound/soc/atmel/sam9g20_wm8731.c +++ b/sound/soc/atmel/sam9g20_wm8731.c @@ -148,7 +148,8 @@ static int at91sam9g20ek_audio_probe(struct platform_device *pdev) codec_np = of_parse_phandle(np, "atmel,audio-codec", 0); if (!codec_np) { dev_err(&pdev->dev, "codec info missing\n"); - return -EINVAL; + ret = -EINVAL; + goto err; } at91sam9g20ek_dai.codecs->of_node = codec_np; @@ -159,7 +160,8 @@ static int at91sam9g20ek_audio_probe(struct platform_device *pdev) if (!cpu_np) { dev_err(&pdev->dev, "dai and pcm info missing\n"); of_node_put(codec_np); - return -EINVAL; + ret = -EINVAL; + goto err; } at91sam9g20ek_dai.cpus->of_node = cpu_np; at91sam9g20ek_dai.platforms->of_node = cpu_np; @@ -170,9 +172,10 @@ static int at91sam9g20ek_audio_probe(struct platform_device *pdev) ret = snd_soc_register_card(card); if (ret) { dev_err(&pdev->dev, "snd_soc_register_card() failed\n"); + goto err; } - return ret; + return 0; err: atmel_ssc_put_audio(0);
The error handling in the AT91SAM9G20-EK machine driver probe did not consistently free the SSC in error paths, sometimes immediately returning an error rather than doing cleanup. Fix this. Signed-off-by: Mark Brown <broonie@kernel.org> --- sound/soc/atmel/sam9g20_wm8731.c | 9 ++++++--- 1 file changed, 6 insertions(+), 3 deletions(-)