2022-07-20 13:14:00

by Charl Liu

Subject: [PATCH V1 1/1] mmc:sdhci:Fix the SD tuning CMD parameter issue that the SDHCI_TRANSFER_MODE is cleared from read direction to write direction incorrectly

When cmd->opcode == MMC_SEND_TUNING_BLOCK, the SDHCI_TRANSFER_MODE
should also be kept

Signed-off-by: Charl Liu <[email protected]>
---
Change in V1:
Keeping the SDHCI_TRANSFER_MODE when cmd->opcode == MMC_END_TUNING_
BLOCK
---
drivers/mmc/host/sdhci.c | 3 ++-
1 file changed, 2 insertions(+), 1 deletion(-)

diff --git a/drivers/mmc/host/sdhci.c b/drivers/mmc/host/sdhci.c
index 22152029e14c..6a0f6725dadb 100644
--- a/drivers/mmc/host/sdhci.c
+++ b/drivers/mmc/host/sdhci.c
@@ -1430,7 +1430,8 @@ 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 ((cmd->opcode != MMC_SEND_TUNING_BLOCK_HS200) &&
+ (cmd->opcode != MMC_SEND_TUNING_BLOCK))
sdhci_writew(host, 0x0, SDHCI_TRANSFER_MODE);
} else {
/* clear Auto CMD settings for no data CMDs */
--
2.32.0

2022-07-24 07:31:23

by Adrian Hunter

Subject: Re: [PATCH V1 1/1] mmc:sdhci:Fix the SD tuning CMD parameter issue that the SDHCI_TRANSFER_MODE is cleared from read direction to write direction incorrectly

On 20/07/22 15:37, Charl Liu wrote:
> When cmd->opcode == MMC_SEND_TUNING_BLOCK, the SDHCI_TRANSFER_MODE
> should also be kept
>
> Signed-off-by: Charl Liu <[email protected]>
> ---
> Change in V1:
> Keeping the SDHCI_TRANSFER_MODE when cmd->opcode == MMC_END_TUNING_
> BLOCK
> ---
> drivers/mmc/host/sdhci.c | 3 ++-
> 1 file changed, 2 insertions(+), 1 deletion(-)
>
> diff --git a/drivers/mmc/host/sdhci.c b/drivers/mmc/host/sdhci.c
> index 22152029e14c..6a0f6725dadb 100644
> --- a/drivers/mmc/host/sdhci.c
> +++ b/drivers/mmc/host/sdhci.c
> @@ -1430,7 +1430,8 @@ 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 ((cmd->opcode != MMC_SEND_TUNING_BLOCK_HS200) &&
> + (cmd->opcode != MMC_SEND_TUNING_BLOCK))

mmc_op_tuning() could be used here

> sdhci_writew(host, 0x0, SDHCI_TRANSFER_MODE);
> } else {
> /* clear Auto CMD settings for no data CMDs */

2022-08-31 13:11:35

by Chevron Li (WH)

Subject: 回复: [PATCH V1 1/1] mmc:sdhci:Fix the SD t uning CMD parameter issue that the SDHCI_TRAN SFER_MODE is cleared from read direction to w rite direction incorrectly

Hi, Adrian,

Sorry response you so late.

Does your comment "mmc_op_tuning() could be used here" means that we should use mmc_send_tuning() instead of sdhci_send_tuning?

If yes, I think we need add below patch commit by Charl Liu for those reasons:
1.Bayhub host tuning command does not have a data payload and our hardware does it automatically, so mmc_send_tuning() will return -EIO.
2.Bayhub host can be used for both SD and eMMC cards.
3.Both MMC_SEND_TUNING_BLOCK and MMC_SEND_TUNING_BLOCK_HS200 are tuning command opcode, maybe the same logical at here is better?

BR,
Chevron

-----邮件原件-----
发件人: Adrian Hunter <[email protected]>
发送时间: Sunday, July 24, 2022 15:22
收件人: Charl Liu (WH) <[email protected]>; [email protected]; [email protected]; [email protected]
抄送: Shaper Liu (WH) <[email protected]>; Chevron Li (WH) <[email protected]>; Thomas Hu(WH) <[email protected]>; XiaoGuang Yu (WH) <[email protected]>; Shirley Her(SC) <[email protected]>
主题: Re: [PATCH V1 1/1] mmc:sdhci:Fix the SD tuning CMD parameter issue that the SDHCI_TRANSFER_MODE is cleared from read direction to write direction incorrectly

On 20/07/22 15:37, Charl Liu wrote:
> When cmd->opcode == MMC_SEND_TUNING_BLOCK, the SDHCI_TRANSFER_MODE
> should also be kept
>
> Signed-off-by: Charl Liu <[email protected]>
> ---
> Change in V1:
> Keeping the SDHCI_TRANSFER_MODE when cmd->opcode == MMC_END_TUNING_
> BLOCK
> ---
> drivers/mmc/host/sdhci.c | 3 ++-
> 1 file changed, 2 insertions(+), 1 deletion(-)
>
> diff --git a/drivers/mmc/host/sdhci.c b/drivers/mmc/host/sdhci.c index
> 22152029e14c..6a0f6725dadb 100644
> --- a/drivers/mmc/host/sdhci.c
> +++ b/drivers/mmc/host/sdhci.c
> @@ -1430,7 +1430,8 @@ 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 ((cmd->opcode != MMC_SEND_TUNING_BLOCK_HS200) &&
> + (cmd->opcode != MMC_SEND_TUNING_BLOCK))

mmc_op_tuning() could be used here

> sdhci_writew(host, 0x0, SDHCI_TRANSFER_MODE);
> } else {
> /* clear Auto CMD settings for no data CMDs */

2022-08-31 14:25:11

by Adrian Hunter

Subject: Re: 回复: [PATCH V1 1/1] mmc:sdhci:Fix t he SD tuning CMD parameter issue that the SDHCI_TRANSFER_MODE is cleared from read direction to write direction incorrectly

On 31/08/22 15:04, Chevron Li (WH) wrote:
> Hi, Adrian,
>
> Sorry response you so late.
>
> Does your comment "mmc_op_tuning() could be used here" means that we should use mmc_send_tuning() instead of sdhci_send_tuning?

I just meant, instead of:

- if (cmd->opcode != MMC_SEND_TUNING_BLOCK_HS200)
+ if ((cmd->opcode != MMC_SEND_TUNING_BLOCK_HS200) &&
+ (cmd->opcode != MMC_SEND_TUNING_BLOCK))

this is simpler:

- if (cmd->opcode != MMC_SEND_TUNING_BLOCK_HS200)
+ if (!mmc_op_tuning(cmd->opcode))


>
> If yes, I think we need add below patch commit by Charl Liu for those reasons:
> 1.Bayhub host tuning command does not have a data payload and our hardware does it automatically, so mmc_send_tuning() will return -EIO.
> 2.Bayhub host can be used for both SD and eMMC cards.
> 3.Both MMC_SEND_TUNING_BLOCK and MMC_SEND_TUNING_BLOCK_HS200 are tuning command opcode, maybe the same logical at here is better?
>
> BR,
> Chevron
>
> -----邮件原件-----
> 发件人: Adrian Hunter <[email protected]>
> 发送时间: Sunday, July 24, 2022 15:22
> 收件人: Charl Liu (WH) <[email protected]>; [email protected]; [email protected]; [email protected]
> 抄送: Shaper Liu (WH) <[email protected]>; Chevron Li (WH) <[email protected]>; Thomas Hu(WH) <[email protected]>; XiaoGuang Yu (WH) <[email protected]>; Shirley Her(SC) <[email protected]>
> 主题: Re: [PATCH V1 1/1] mmc:sdhci:Fix the SD tuning CMD parameter issue that the SDHCI_TRANSFER_MODE is cleared from read direction to write direction incorrectly
>
> On 20/07/22 15:37, Charl Liu wrote:
>> When cmd->opcode == MMC_SEND_TUNING_BLOCK, the SDHCI_TRANSFER_MODE
>> should also be kept
>>
>> Signed-off-by: Charl Liu <[email protected]>
>> ---
>> Change in V1:
>> Keeping the SDHCI_TRANSFER_MODE when cmd->opcode == MMC_END_TUNING_
>> BLOCK
>> ---
>> drivers/mmc/host/sdhci.c | 3 ++-
>> 1 file changed, 2 insertions(+), 1 deletion(-)
>>
>> diff --git a/drivers/mmc/host/sdhci.c b/drivers/mmc/host/sdhci.c index
>> 22152029e14c..6a0f6725dadb 100644
>> --- a/drivers/mmc/host/sdhci.c
>> +++ b/drivers/mmc/host/sdhci.c
>> @@ -1430,7 +1430,8 @@ 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 ((cmd->opcode != MMC_SEND_TUNING_BLOCK_HS200) &&
>> + (cmd->opcode != MMC_SEND_TUNING_BLOCK))
>
> mmc_op_tuning() could be used here
>
>> sdhci_writew(host, 0x0, SDHCI_TRANSFER_MODE);
>> } else {
>> /* clear Auto CMD settings for no data CMDs */
>