[PATCH v3 12/25] mtd: spi-nor: core: Call spi_nor_post_sfdp_fixups() only when SFDP is defined

[Michael Walle]

Am 2021-10-29 19:26, schrieb Tudor Ambarus:
> spi_nor_post_sfdp_fixups() was called even when there were no SFDP
> tables defined. late_init() should be instead used for flashes that
> do not define SFDP tables.
> 
> Use spi_nor_post_sfdp_fixups() just to fix SFDP data. post_sfdp()
> hook is as of now used just by s28hs512t, mt35xu512aba, and both
> support SFDP, there's no functional change with this patch.
> 
> Signed-off-by: Tudor Ambarus <tudor.ambarus at microchip.com>
> ---
>  drivers/mtd/spi-nor/core.c | 56 ++++++++++++++++++--------------------
>  1 file changed, 26 insertions(+), 30 deletions(-)
> 
> diff --git a/drivers/mtd/spi-nor/core.c b/drivers/mtd/spi-nor/core.c
> index 4679298c5301..82cc56c9d09e 100644
> --- a/drivers/mtd/spi-nor/core.c
> +++ b/drivers/mtd/spi-nor/core.c
> @@ -2508,6 +2508,25 @@ static void
> spi_nor_manufacturer_init_params(struct spi_nor *nor)
>  		nor->info->fixups->default_init(nor);
>  }
> 
> +/**
> + * spi_nor_post_sfdp_fixups() - Updates the flash's parameters and 
> settings
> + * after SFDP has been parsed. Called only for flashes that define 
> JESD216 SFDP
> + * tables.
> + * @nor:	pointer to a 'struct spi_nor'
> + *
> + * Used to tweak various flash parameters when information provided by 
> the SFDP
> + * tables are wrong.
> + */
> +static void spi_nor_post_sfdp_fixups(struct spi_nor *nor)
> +{
> +	if (nor->manufacturer && nor->manufacturer->fixups &&
> +	    nor->manufacturer->fixups->post_sfdp)
> +		nor->manufacturer->fixups->post_sfdp(nor);
> +
> +	if (nor->info->fixups && nor->info->fixups->post_sfdp)
> +		nor->info->fixups->post_sfdp(nor);
> +}
> +
>  /**
>   * spi_nor_sfdp_init_params() - Initialize the flash's parameters and 
> settings
>   * based on JESD216 SFDP standard.
> @@ -2522,7 +2541,9 @@ static void spi_nor_sfdp_init_params(struct 
> spi_nor *nor)
> 
>  	memcpy(&sfdp_params, nor->params, sizeof(sfdp_params));
> 
> -	if (spi_nor_parse_sfdp(nor)) {
> +	if (!spi_nor_parse_sfdp(nor)) {
> +		spi_nor_post_sfdp_fixups(nor);

I find this function particulary confusing. Why is it
copying the params around? I know the function doc
says rollback, but can't we make this better?

Either make spi_nor_parse_sfdp() commit the nor->params update
atomically, or pass a second parameter sfdp_params, which are
copied to nor->params here if spi_nor_parse_sfdp() was successful.

Also the control flow could be better.

ret = spi_nor_parse_sfdp(nor, &sfdp_params);
if (!ret) {
     /* clever comment, why is addr_width = 0 here */
     nor->addr_width = 0;
     nor->flags &= ~SNOR_F_4B_OPCODES;
     return 0;
}

memcpy(nor->params, &sfdp_params, sizeof(*nor->params));
spi_nor_post_sfdp_fixups(nor);

Having an even closer look, addr_width is set to 0 because
spi_nor_parse_sfdp() is also changing that. nor->flags
is also changed and not only the SNOR_F_4B_OPCODES bit. But
only that one is cleared?!

I think this deserves another cleanup series. Having the
rollback here makes no sense. You'd have to keep both in
sync.

IMHO best would be a second parameter and make sure all
code in sfdp.c don't change anything else. Which would
probably mean that addr_width and flags will move to
nor->params.

Anyway, for now:
Reviewed-by: Michael Walle <michael at walle.cc>

-michael