Hi,
> [PATCH 2/4] misc: eeprom_93xx46: set size and addrlen according to the dts
This patch doesn't really deal with the devicetree, so this subject line
seems a bit mismatched.
On Sat, Apr 24, 2021 at 02:30:31PM +0200, Emmanuel Gil Peyrot wrote:
> This can then be used by the rest of the driver to use the correct
> commands on 93c56 and 93c66.
>
> Signed-off-by: Emmanuel Gil Peyrot <linkmauve@xxxxxxxxxxxx>
> ---
Ah hmmm. Does this mean that with the previous patch, the driver will be
instanciated for 93c56 and 93c66 but send the wrong commands? I think
you should avoid this pitfall by rearranging (or squashing) the patches.
> drivers/misc/eeprom/eeprom_93xx46.c | 21 +++++++++++++++++----
> 1 file changed, 17 insertions(+), 4 deletions(-)
>
> diff --git a/drivers/misc/eeprom/eeprom_93xx46.c b/drivers/misc/eeprom/eeprom_93xx46.c
> index 64dd76f66463..39375255e22a 100644
> --- a/drivers/misc/eeprom/eeprom_93xx46.c
> +++ b/drivers/misc/eeprom/eeprom_93xx46.c
> @@ -18,6 +18,7 @@
> #include <linux/spi/spi.h>
> #include <linux/nvmem-provider.h>
> #include <linux/eeprom_93xx46.h>
> +#include <linux/log2.h>
>
> #define OP_START 0x4
> #define OP_WRITE (OP_START | 0x1)
> @@ -474,10 +475,22 @@ static int eeprom_93xx46_probe(struct spi_device *spi)
> if (!edev)
> return -ENOMEM;
>
> + if (pd->flags & EE_SIZE1K)
> + edev->size = 128;
> + else if (pd->flags & EE_SIZE2K)
> + edev->size = 256;
> + else if (pd->flags & EE_SIZE4K)
> + edev->size = 512;
> + else {
> + dev_err(&spi->dev, "unspecified size\n");
> + err = -EINVAL;
> + goto fail;
> + }
> +
> if (pd->flags & EE_ADDR8)
> - edev->addrlen = 7;
> + edev->addrlen = ilog2(edev->size);
> else if (pd->flags & EE_ADDR16)
> - edev->addrlen = 6;
> + edev->addrlen = ilog2(edev->size) - 1;
> else {
> dev_err(&spi->dev, "unspecified address type\n");
> return -EINVAL;
> @@ -488,7 +501,6 @@ static int eeprom_93xx46_probe(struct spi_device *spi)
> edev->spi = spi;
> edev->pdata = pd;
>
> - edev->size = 128;
> edev->nvmem_config.type = NVMEM_TYPE_EEPROM;
> edev->nvmem_config.name = dev_name(&spi->dev);
> edev->nvmem_config.dev = &spi->dev;
> @@ -508,8 +520,9 @@ static int eeprom_93xx46_probe(struct spi_device *spi)
> if (IS_ERR(edev->nvmem))
> return PTR_ERR(edev->nvmem);
>
> - dev_info(&spi->dev, "%d-bit eeprom %s\n",
> + dev_info(&spi->dev, "%d-bit eeprom containing %d bytes %s\n",
> (pd->flags & EE_ADDR8) ? 8 : 16,
> + edev->size,
> (pd->flags & EE_READONLY) ? "(readonly)" : "");
The logic itself looks good though.
Thanks,
Jonathan
Attachment:
signature.asc
Description: PGP signature
