Re: [PATCH V2 1/1] mmc:sdhci:Fix the SD tuning issue that the SDHCI_TRANSFER_MODE is cleared incorrectly

[Adrian Hunter]

On 4/11/22 08:42, Charl Liu wrote:
> When cmd->opcode == MMC_SEND_TUNING_BLOCK, the SDHCI_TRANSFER_MODE
> should also be kept
> 
> Signed-off-by: Charl Liu <charl.liu@xxxxxxxxxxxxxx>
> ---
> change in V1:
> Keeping the SDHCI_TRANSFER_MODE when cmd->opcode == MMC_END_TUNING_BLOCK
> 
> change in V2:
> add the mmc_op_tuning interface to judge if the opcode is tuning CMD
> ---
>  drivers/mmc/host/sdhci.c | 2 +-
>  include/linux/mmc/mmc.h  | 6 ++++++
>  2 files changed, 7 insertions(+), 1 deletion(-)
> 
> diff --git a/drivers/mmc/host/sdhci.c b/drivers/mmc/host/sdhci.c
> index 22152029e14c..9388628215ca 100644
> --- a/drivers/mmc/host/sdhci.c
> +++ b/drivers/mmc/host/sdhci.c
> @@ -1430,7 +1430,7 @@ static void sdhci_set_transfer_mode(struct sdhci_host *host,
>  		if (host->quirks2 &
>  			SDHCI_QUIRK2_CLEAR_TRANSFERMODE_REG_BEFORE_CMD) {
>  			/* must not clear SDHCI_TRANSFER_MODE when tuning */
> -			if (cmd->opcode != MMC_SEND_TUNING_BLOCK_HS200)
> +			if (!mmc_op_tuning(cmd->opcode))
>  				sdhci_writew(host, 0x0, SDHCI_TRANSFER_MODE);
>  		} else {
>  		/* clear Auto CMD settings for no data CMDs */
> diff --git a/include/linux/mmc/mmc.h b/include/linux/mmc/mmc.h
> index d9a65c6a8816..d809bdfcc59b 100644
> --- a/include/linux/mmc/mmc.h
> +++ b/include/linux/mmc/mmc.h
> @@ -99,6 +99,12 @@ static inline bool mmc_op_multi(u32 opcode)
>  	       opcode == MMC_READ_MULTIPLE_BLOCK;
>  }
>  
> +static inline bool mmc_op_tuning(u32 opcode)
> +{
> +	return opcode == MMC_SEND_TUNING_BLOCK ||
> +	       opcode == MMC_SEND_TUNING_BLOCK_HS200;
> +}
> +

Patches for mmc should be based on Ulf's next branch:

https://git.kernel.org/pub/scm/linux/kernel/git/ulfh/mmc.git/log/?h=next

which already has mmc_op_tuning()

>  /*
>   * MMC_SWITCH argument format:
>   *