On 14/11/22 15:26, Christian Löhle wrote:
> Discard feature is independent of security features.
> The support check for all trims and discard falsely checked
> for secure trim/discard, but in the discard case this is not
> mandated by the spec.
>
"Discard" was added after "Trim", so while MMC_TRIM_ARGS made
sense originally, that stopped with:
commit b3bf915308ca ("mmc: core: new discard feature support at eMMC v4.5")
So this could have that as a fixes tag.
Also MMC_TRIM_ARGS is a bit of a misleading name since it covers
one of the bits of MMC_DISCARD_ARG, so I suggest a second patch to
tidy things up.
Perhaps rename it and change the value to 8003 e.g.
#define MMC_TRIM_OR_DISCARD_ARGS 0x00008003
and make a wrapper for the Trim case:
static bool is_trim_arg(unsigned int arg)
{
return (arg & MMC_TRIM_OR_DISCARD_ARGS) && arg != MMC_DISCARD_ARG;
}
> Signed-off-by: Christian Loehle <cloehle@xxxxxxxxxxxxxx>
> ---
> drivers/mmc/core/core.c | 3 ++-
> 1 file changed, 2 insertions(+), 1 deletion(-)
>
> diff --git a/drivers/mmc/core/core.c b/drivers/mmc/core/core.c
> index 95fa8fb1d45f..507005211529 100644
> --- a/drivers/mmc/core/core.c
> +++ b/drivers/mmc/core/core.c
> @@ -1761,7 +1761,8 @@ int mmc_erase(struct mmc_card *card, unsigned int from, unsigned int nr,
> return -EOPNOTSUPP;
>
> if (mmc_card_mmc(card) && (arg & MMC_TRIM_ARGS) &&
> - !(card->ext_csd.sec_feature_support & EXT_CSD_SEC_GB_CL_EN))
> + !(card->ext_csd.sec_feature_support & EXT_CSD_SEC_GB_CL_EN) &&
> + arg != MMC_DISCARD_ARG)
> return -EOPNOTSUPP;
>
> if (arg == MMC_SECURE_ERASE_ARG) {
