Re: [PATCH 03/14] mmc: mmci: Add support for setting pad type via pinctrl

[Ulf Hansson]

On 12 January 2018 at 13:15,  <patrice.chotard@xxxxxx> wrote:
> From: Patrice Chotard <patrice.chotard@xxxxxx>
>
> The STM32 variant hasn't the control bit to switch pads in opendrain mode.
> In this case we can achieve the same result by asking to the pinmux driver
> to configure pins for us.
>
> This patch make the mmci driver able to do this whenever needed.
>
> Signed-off-by: Andrea Merello <andrea.merello@xxxxxxxxx>
> Signed-off-by: Patrice Chotard <patrice.chotard@xxxxxx>
> ---
>  drivers/mmc/host/mmci.c | 54 ++++++++++++++++++++++++++++++++++++++++---------
>  drivers/mmc/host/mmci.h |  5 +++++
>  2 files changed, 50 insertions(+), 9 deletions(-)
>
> diff --git a/drivers/mmc/host/mmci.c b/drivers/mmc/host/mmci.c
> index 7e56f85..38e8c20 100644
> --- a/drivers/mmc/host/mmci.c
> +++ b/drivers/mmc/host/mmci.c
> @@ -85,6 +85,8 @@
>   * @mmcimask1: true if variant have a MMCIMASK1 register.
>   * @start_err: true is the variant has STARTBITERR bit inside MMCISTATUS
>   *            register.
> + * @opendrain: true if variant have dedicated bit for opendrain pins
> + *            configuration.
>   */
>  struct variant_data {
>         unsigned int            clkreg;
> @@ -116,6 +118,7 @@ struct variant_data {
>         bool                    reversed_irq_handling;
>         bool                    mmcimask1;
>         bool                    start_err;
> +       bool                    opendrain;

Similar comment as for patch2.

To be consistent with how we implement support for similar variant
variations, I would prefer to have this being a u32. Something along
the lines of how the "busy_detect_flag" is being used.

[...]

> @@ -1394,9 +1405,11 @@ static void mmci_set_ios(struct mmc_host *mmc, struct mmc_ios *ios)
>  {
>         struct mmci_host *host = mmc_priv(mmc);
>         struct variant_data *variant = host->variant;
> +       struct pinctrl_state *pins;
>         u32 pwr = 0;
>         unsigned long flags;
>         int ret;
> +       bool is_opendrain;
>
>         if (host->plat->ios_handler &&
>                 host->plat->ios_handler(mmc_dev(mmc), ios))
> @@ -1455,16 +1468,31 @@ static void mmci_set_ios(struct mmc_host *mmc, struct mmc_ios *ios)
>                                 ~MCI_ST_DATA2DIREN);
>         }
>
> -       if (ios->bus_mode == MMC_BUSMODE_OPENDRAIN) {
> -               if (host->hw_designer != AMBA_VENDOR_ST)
> -                       pwr |= MCI_ROD;
> -               else {
> -                       /*
> -                        * The ST Micro variant use the ROD bit for something
> -                        * else and only has OD (Open Drain).
> -                        */
> -                       pwr |= MCI_OD;

Seems like you should actually split this change into two parts.

One that adds the variant flag for the open drain bit, when then can
clean up this code. Then a patch on top that starts using pinctrl in
case there is no open drain bit set.

Does that sounds reasonable?

> +       if (host->variant->opendrain) {
> +               if (ios->bus_mode == MMC_BUSMODE_OPENDRAIN) {
> +                       if (host->hw_designer != AMBA_VENDOR_ST) {
> +                               pwr |= MCI_ROD;
> +                       } else {
> +                               /*
> +                                * The ST Micro variant use the ROD bit for
> +                                * something else and only has OD (Open Drain).
> +                                */
> +                               pwr |= MCI_OD;
> +                       }
>                 }
> +       } else {
> +               /*
> +                * If the variant cannot configure the pads by its own, then we
> +                * expect the pinctrl to be able to do that for us
> +                */
> +               is_opendrain = (ios->bus_mode == MMC_BUSMODE_OPENDRAIN);
> +               pins = pinctrl_lookup_state(host->pinctrl, is_opendrain ?

How about doing the lookup in ->probe() instead? Then just select the
state here, if supported?

> +                                       MMCI_PINCTRL_STATE_OPENDRAIN :
> +                                       MMCI_PINCTRL_STATE_PUSHPULL);
> +               if (IS_ERR(pins))
> +                       dev_warn(mmc_dev(mmc), "Cannot select pin drive type via pinctrl\n");
> +               else
> +                       pinctrl_select_state(host->pinctrl, pins);
>         }
>
>         /*
> @@ -1609,6 +1637,14 @@ static int mmci_probe(struct amba_device *dev,
>         host = mmc_priv(mmc);
>         host->mmc = mmc;
>
> +       if (!variant->opendrain) {
> +               host->pinctrl = devm_pinctrl_get(&dev->dev);
> +               if (IS_ERR(host->pinctrl)) {
> +                       dev_err(&dev->dev, "failed to get pinctrl");
> +                       goto host_free;
> +               }
> +       }
> +
>         host->hw_designer = amba_manf(dev);
>         host->hw_revision = amba_rev(dev);
>         dev_dbg(mmc_dev(mmc), "designer ID = 0x%02x\n", host->hw_designer);
> diff --git a/drivers/mmc/host/mmci.h b/drivers/mmc/host/mmci.h
> index 83160a9..de3d0b3 100644
> --- a/drivers/mmc/host/mmci.h
> +++ b/drivers/mmc/host/mmci.h
> @@ -192,6 +192,10 @@
>
>  #define NR_SG          128
>
> +/* pinctrl configs */
> +#define MMCI_PINCTRL_STATE_PUSHPULL "default"

Seems like we should be able to cope fine without having to add a
separate define for the PUSHPULL, but rather just select the default
state instead.

> +#define MMCI_PINCTRL_STATE_OPENDRAIN "opendrain"
> +
>  struct clk;
>  struct variant_data;
>  struct dma_chan;
> @@ -227,6 +231,7 @@ struct mmci_host {
>         bool                    vqmmc_enabled;
>         struct mmci_platform_data *plat;
>         struct variant_data     *variant;
> +       struct pinctrl          *pinctrl;
>
>         u8                      hw_designer;
>         u8                      hw_revision:4;
> --
> 1.9.1
>

Kind regards
Uffe