[PATCH v2] iio: adc: ad799x: Add a single error handling block at the end of the function.

[PATCH v2] iio: adc: ad799x: Add a single error handling block at the end of the function.

[Cai Huoqing]

A single error handling block at the end of the function could
be brought in to make code a little more concise.

Signed-off-by: Cai Huoqing <caihuoqing@baidu.com>
---
v1->v2:
Change the error path-error_disable_reg from "st->vref" to "st->reg".

 drivers/iio/adc/ad799x.c | 17 ++++++++++-------
 1 file changed, 10 insertions(+), 7 deletions(-)

diff --git a/drivers/iio/adc/ad799x.c b/drivers/iio/adc/ad799x.c
index 18bf8386d50a..d3dbc4c1e375 100644
--- a/drivers/iio/adc/ad799x.c
+++ b/drivers/iio/adc/ad799x.c
@@ -891,20 +891,23 @@ static int __maybe_unused ad799x_resume(struct device *dev)
 	}
 	ret = regulator_enable(st->vref);
 	if (ret) {
-		regulator_disable(st->reg);
 		dev_err(dev, "Unable to enable vref regulator\n");
-		return ret;
+		goto error_disable_reg;
 	}
 
 	/* resync config */
 	ret = ad799x_update_config(st, st->config);
-	if (ret) {
-		regulator_disable(st->vref);
-		regulator_disable(st->reg);
-		return ret;
-	}
+	if (ret)
+		goto error_disable_vref;
 
 	return 0;
+
+error_disable_vref:
+	regulator_disable(st->vref);
+error_disable_reg:
+	regulator_disable(st->reg);
+
+	return ret;
 }
 
 static SIMPLE_DEV_PM_OPS(ad799x_pm_ops, ad799x_suspend, ad799x_resume);
-- 
2.25.1

Re: [PATCH v2] iio: adc: ad799x: Add a single error handling block at the end of the function.

[Jonathan Cameron]

On Sun, 12 Sep 2021 14:41:01 +0800
Cai Huoqing <caihuoqing@baidu.com> wrote:

> A single error handling block at the end of the function could
> be brought in to make code a little more concise.
> 
> Signed-off-by: Cai Huoqing <caihuoqing@baidu.com>
+CC Florian who is reworking some of the regulator handling in this driver.

https://lore.kernel.org/linux-iio/20210908152525.2946785-1-florian.boor@kernelconcepts.de/

One suggestion I made to Florian's v2 was to make the enabling of st->vref
only occur it is actually present (rather than no there after making it
optional).  That is going to cause merge problems against this patch.

@Florian, would you mind adding Cai Huoqing's patch to your series so that
we get both your improvements and his on one go without conflicts?
Either apply it as a precursor or rebase it on top of your patches.

Thanks,

Jonathan

> ---
> v1->v2:
> Change the error path-error_disable_reg from "st->vref" to "st->reg".
> 
>  drivers/iio/adc/ad799x.c | 17 ++++++++++-------
>  1 file changed, 10 insertions(+), 7 deletions(-)
> 
> diff --git a/drivers/iio/adc/ad799x.c b/drivers/iio/adc/ad799x.c
> index 18bf8386d50a..d3dbc4c1e375 100644
> --- a/drivers/iio/adc/ad799x.c
> +++ b/drivers/iio/adc/ad799x.c
> @@ -891,20 +891,23 @@ static int __maybe_unused ad799x_resume(struct device *dev)
>  	}
>  	ret = regulator_enable(st->vref);
>  	if (ret) {
> -		regulator_disable(st->reg);
>  		dev_err(dev, "Unable to enable vref regulator\n");
> -		return ret;
> +		goto error_disable_reg;
>  	}
>  
>  	/* resync config */
>  	ret = ad799x_update_config(st, st->config);
> -	if (ret) {
> -		regulator_disable(st->vref);
> -		regulator_disable(st->reg);
> -		return ret;
> -	}
> +	if (ret)
> +		goto error_disable_vref;
>  
>  	return 0;
> +
> +error_disable_vref:
> +	regulator_disable(st->vref);
> +error_disable_reg:
> +	regulator_disable(st->reg);
> +
> +	return ret;
>  }
>  
>  static SIMPLE_DEV_PM_OPS(ad799x_pm_ops, ad799x_suspend, ad799x_resume);