Re: [PATCH V12 12/23] mmc: sdhci-uhs2: skip signal_voltage_switch()

From: Adrian Hunter
Date: Fri Oct 06 2023 - 06:51:18 EST


On 6/10/23 13:30, Victor Shih wrote:
> On Tue, Oct 3, 2023 at 5:58 PM Ulf Hansson <ulf.hansson@xxxxxxxxxx> wrote:
>>
>> On Fri, 15 Sept 2023 at 11:44, Victor Shih <victorshihgli@xxxxxxxxx> wrote:
>>>
>>> From: Victor Shih <victor.shih@xxxxxxxxxxxxxxxxxxx>
>>>
>>> For UHS2, the signal voltage is supplied by vdd2 which is already 1.8v,
>>> so no voltage switch required.
>>
>> Can you please elaborate on this? I don't get anything of the above, sorry.
>>
>>>
>>> Signed-off-by: Ben Chuang <ben.chuang@xxxxxxxxxxxxxxxxxxx>
>>> Signed-off-by: AKASHI Takahiro <takahiro.akashi@xxxxxxxxxx>
>>> Signed-off-by: Victor Shih <victor.shih@xxxxxxxxxxxxxxxxxxx>
>>> Acked-by: Adrian Hunter <adrian.hunter@xxxxxxxxx>
>>> ---
>>>
>>> Updates in V5:
>>> - Use sdhci_uhs2_mode() to simplify code in
>>> sdhci_uhs2_start_signal_voltage_switch().
>>>
>>> ---
>>>
>>> drivers/mmc/host/sdhci-uhs2.c | 24 ++++++++++++++++++++++++
>>> 1 file changed, 24 insertions(+)
>>>
>>> diff --git a/drivers/mmc/host/sdhci-uhs2.c b/drivers/mmc/host/sdhci-uhs2.c
>>> index fc37a34629c2..92fb69b7e209 100644
>>> --- a/drivers/mmc/host/sdhci-uhs2.c
>>> +++ b/drivers/mmc/host/sdhci-uhs2.c
>>> @@ -142,6 +142,27 @@ static void sdhci_uhs2_set_power(struct sdhci_host *host, unsigned char mode, un
>>> }
>>> }
>>>
>>> +/*****************************************************************************\
>>> + * *
>>> + * MMC callbacks *
>>> + * *
>>> +\*****************************************************************************/
>>> +
>>> +static int sdhci_uhs2_start_signal_voltage_switch(struct mmc_host *mmc,
>>> + struct mmc_ios *ios)
>>> +{
>>> + struct sdhci_host *host = mmc_priv(mmc);
>>> +
>>> + /*
>>> + * For UHS2, the signal voltage is supplied by vdd2 which is
>>> + * already 1.8v so no voltage switch required.
>>> + */
>>> + if (sdhci_uhs2_mode(host))
>>> + return 0;
>>
>> This is just wrong. If we are initializing a uhs2 card, we certainly
>> should call ->start_signal_voltage_switch() callback at all. This is
>> for UHS-I cards, right?
>>
>
> Hi, Ulf
>
> UHS-II does not need single_voltage.
> I will modify the commit message in the next version.
> sdhci_uhs2_start_signal_voltage_switch() is under
> mmc_host_ops.start_signal_voltage_switch host ops,
> therefore, we need to keep the UHS-I path here.

You should be able to leave out the patch entirely
because ->start_signal_voltage_switch() is not called
for UHS2 mode

>
> Thanks, Victor Shih
>
>>> +
>>> + return sdhci_start_signal_voltage_switch(mmc, ios);
>>> +}
>>> +
>>> /*****************************************************************************\
>>> * *
>>> * Driver init/exit *
>>> @@ -150,6 +171,9 @@ static void sdhci_uhs2_set_power(struct sdhci_host *host, unsigned char mode, un
>>>
>>> static int sdhci_uhs2_host_ops_init(struct sdhci_host *host)
>>> {
>>> + host->mmc_host_ops.start_signal_voltage_switch =
>>> + sdhci_uhs2_start_signal_voltage_switch;
>>> +
>>> return 0;
>>> }
>>>
>>
>> Kind regards
>> Uffe