Re: [PATCH] iio: light: tsl2591: reduce code duplication in tsl2591_persist_* functions

[David Lechner <dlechner@baylibre.com>]

On 4/18/26 1:54 PM, LucasRabaquim wrote:
> Add tsl2591_persist_table as a bidirectional lookup table for
> tsl2591_persist_cycle_to_lit and tsl2591_persist_lit_to_cycle functions
> 
> These functions were prone to errors in the case of adding or updating
> TSL2591_PRST_ALS_INT_CYCLE_* values. The functions utilize a for-loop

Isn't the new table also prone to errors? Neither is very complicated.

Seems like the advantage is rather the reduction in verbosity.

> like in the function tsl2591_als_time_to_fval, allowing it to compare
> the cycle value with tsl2591_persist_table[i].cycle and return lit
> value with tsl2591_persist_table[i].lit as defined in the new type
> tsl2591_persist_entry, reducing the risk of logic mismatches.
> 
> Signed-off-by: LucasRabaquim <lucas.rabaquim@usp.br>
> Co-developed-by: Matheus Feitosa <matheus.feitosa@usp.br>
> Signed-off-by: Matheus Feitosa <matheus.feitosa@usp.br>
> ---
>  drivers/iio/light/tsl2591.c | 104 +++++++++++++-----------------------
>  1 file changed, 38 insertions(+), 66 deletions(-)
> 
> diff --git a/drivers/iio/light/tsl2591.c b/drivers/iio/light/tsl2591.c
> index c5557867ea43..86483f0b2e01 100644
> --- a/drivers/iio/light/tsl2591.c
> +++ b/drivers/iio/light/tsl2591.c
> @@ -170,6 +170,12 @@ struct tsl2591_chip {
>  	bool events_enabled;
>  };
>  
> +/* Persist cycle conversion table mapping register cycles to lit values */
> +struct tsl2591_persist_entry {
> +	u8 cycle;
> +	int lit;

Why not u8 for both?

> +};
> +