Re: [PATCH v3 2/2] iio: adc: vf610_adc: limit i.MX6SX's channel number to 4

[Jonathan Cameron <jic23@kernel.org>]

On Thu,  7 Nov 2024 14:18:41 -0500
Frank Li <Frank.Li@nxp.com> wrote:

> i.MX6SX only has 4 ADC channels, so limit channel numbers to 4 for
> compatible string 'fsl,imx6sx-adc'.
> 
> Reviewed-by: Haibo Chen <haibo.chen@nxp.com>
> Signed-off-by: Frank Li <Frank.Li@nxp.com>

I'm pretty strongly against putting integers in the id match
data variables.  Instead use a much more easily extended structure
(which will have one element for now)

struct vf610_chip_info {
	u8 num_channels;
};

and then pointers to that structure in the tables.

The integer path regularly goes wrong (0 causes trouble with error handling)
and so I push back on people doing this in new code.

One other comment inline.

Jonathan

> ---
> compatible string 'fsl,imx6sx-adc' already document in
> Documentation/devicetree/bindings/iio/adc/fsl,vf610-adc.yaml
> 
> Change from v2 to v3
> - none
> 
> Change from v1 to v2
> - Add Haibo Chen <haibo.chen@nxp.com>
> - change cast to uintptr_t to fix below warning
> 
> | Reported-by: kernel test robot <lkp@intel.com>
> | Closes: https://lore.kernel.org/oe-kbuild-all/202411052345.gyJaM3h4-lkp@intel.com/
> All warnings (new ones prefixed by >>):
> 
>    In file included from drivers/iio/adc/vf610_adc.c:20:   In file included from include/linux/regulator/consumer.h:35:
>    In file included from include/linux/suspend.h:5:
>    In file included from include/linux/swap.h:9:
>    In file included from include/linux/memcontrol.h:21:
>    In file included from include/linux/mm.h:2213:
>    include/linux/vmstat.h:518:36: warning: arithmetic between different enumeration types ('enum node_stat_item' and 'enum lru_list') [-Wenum-enum-conversion]
>      518 |         return node_stat_name(NR_LRU_BASE + lru) + 3; // skip "nr_"
>          |                               ~~~~~~~~~~~ ^ ~~~>> drivers/iio/adc/vf610_adc.c:874:28: warning: cast to smaller integer type 'u32' (aka 'unsigned int') from 'const void *' [-Wvoid-pointer-to-int-cast]
>      874 |         indio_dev->num_channels = (u32)device_get_match_data(dev);
>          |
> ---
>  drivers/iio/adc/vf610_adc.c | 5 +++--
>  1 file changed, 3 insertions(+), 2 deletions(-)
> 
> diff --git a/drivers/iio/adc/vf610_adc.c b/drivers/iio/adc/vf610_adc.c
> index a6a0ada8a102f..36f6132bf5ba4 100644
> --- a/drivers/iio/adc/vf610_adc.c
> +++ b/drivers/iio/adc/vf610_adc.c
> @@ -809,7 +809,8 @@ static const struct iio_info vf610_adc_iio_info = {
>  };
>  
>  static const struct of_device_id vf610_adc_match[] = {
> -	{ .compatible = "fsl,vf610-adc", },
> +	{ .compatible = "fsl,imx6sx-adc", .data = (void *)4},
> +	{ .compatible = "fsl,vf610-adc", .data = (void *)ARRAY_SIZE(vf610_adc_iio_channels)},
>  	{ /* sentinel */ }
>  };
>  MODULE_DEVICE_TABLE(of, vf610_adc_match);
> @@ -870,7 +871,7 @@ static int vf610_adc_probe(struct platform_device *pdev)
>  	indio_dev->info = &vf610_adc_iio_info;
>  	indio_dev->modes = INDIO_DIRECT_MODE;
>  	indio_dev->channels = vf610_adc_iio_channels;
> -	indio_dev->num_channels = ARRAY_SIZE(vf610_adc_iio_channels);
> +	indio_dev->num_channels = (uintptr_t)device_get_match_data(dev);
That can in theory fail and return NULL. Which here would be interpreted
as 0 channels and probably cause trouble later.

>  
>  	vf610_adc_cfg_init(info);
>  	vf610_adc_hw_init(info);