Re: [PATCH v3 2/3] iio: ssp_sensors: simplify cleanup using __free

[Jonathan Cameron]

On Sun, 15 Mar 2026 18:25:08 +0530
Sanjay Chitroda <sanjayembeddedse@xxxxxxxxx> wrote:

> From: Sanjay Chitroda <sanjayembeddedse@xxxxxxxxx>
> 
> Replace manual cleanup logic with __free attribute from cleanup.h. This
> removes explicit kfree() calls and simplifies the error handling paths.
> 
> No functional change intended for kmalloc().
> 
> Signed-off-by: Sanjay Chitroda <sanjayembeddedse@xxxxxxxxx>
Hi Sanjay

I'd suggest slowing down with new versions. 1 weeks is a usual good minimum
because it avoids what has happened here where you get feedback on a later
version that could have been given on the earlier one and hence
waste both your time and reviewer time (a little).

> ---
> Changes in v3:
> - prepare series to have all respective cleanup API support for the ssp_sensors following input from Andy Shevchenko
> - Link to v2 https://lore.kernel.org/all/20260311174151.3441429-1-sanjayembedded@xxxxxxxxx/
> Changes in v2:
> - split series to individual patch
> - address review comment from Andy Shevchenko
> - Link to v1 https://lore.kernel.org/all/20260310200513.2162018-3-sanjayembedded@xxxxxxxxx/
> ---
>  drivers/iio/common/ssp_sensors/ssp_spi.c | 14 +++-----------
>  1 file changed, 3 insertions(+), 11 deletions(-)
> 
> diff --git a/drivers/iio/common/ssp_sensors/ssp_spi.c b/drivers/iio/common/ssp_sensors/ssp_spi.c
> index 87a38002b218..9df94bd02e3a 100644
> --- a/drivers/iio/common/ssp_sensors/ssp_spi.c
> +++ b/drivers/iio/common/ssp_sensors/ssp_spi.c
> @@ -326,7 +326,6 @@ static int ssp_parse_dataframe(struct ssp_data *data, char *dataframe, int len)
>  /* threaded irq */
>  int ssp_irq_msg(struct ssp_data *data)
>  {
> -	char *buffer;
>  	u8 msg_type;
>  	int ret;
>  	u16 length, msg_options;
> @@ -370,7 +369,7 @@ int ssp_irq_msg(struct ssp_data *data)
>  			 * but the slave should not send such ones - it is to
>  			 * check but let's handle this
>  			 */
> -			buffer = kmalloc(length, GFP_KERNEL | GFP_DMA);
> +			char *buffer __free(kfree) = kmalloc(length, GFP_KERNEL | GFP_DMA);
Andy raised if this is the right thing to do at all I think.
Maybe look at whether there is a simple maximum length and stash a buffer that
gets reused each time instead of a fresh allocation. May well simplify
the code.

>  			if (!buffer)
>  				return -ENOMEM;
>  
> @@ -379,8 +378,6 @@ int ssp_irq_msg(struct ssp_data *data)
>  			if (ret >= 0)
>  				ret = -EPROTO;
>  
> -			kfree(buffer);
> -
>  			dev_err(SSP_DEV, "No match error %x\n",
>  				msg_options);
>  
> @@ -411,22 +408,17 @@ int ssp_irq_msg(struct ssp_data *data)
>  				complete(msg->done);
>  		break;
>  	case SSP_HUB2AP_WRITE:
> -		buffer = kzalloc(length, GFP_KERNEL | GFP_DMA);
> +		char *buffer __free(kfree) = kzalloc(length, GFP_KERNEL | GFP_DMA);

Scope isn't what you think it is...


>  		if (!buffer)
>  			return -ENOMEM;
>  
>  		ret = spi_read(data->spi, buffer, length);
>  		if (ret < 0) {
>  			dev_err(SSP_DEV, "spi read fail\n");
> -			kfree(buffer);
>  			break;
>  		}
>  
> -		ret = ssp_parse_dataframe(data, buffer, length);
> -
> -		kfree(buffer);
> -		break;
> -
> +		return ssp_parse_dataframe(data, buffer, length);
>  	default:
>  		dev_err(SSP_DEV, "unknown msg type\n");
>  		return -EPROTO;