Re: [PATCH v3 1/2] pinctrl: qcom: Add support for i2c specific pull feature

From: Bjorn Andersson
Date: Wed Feb 08 2023 - 22:45:58 EST


On Wed, Feb 08, 2023 at 10:18:35AM +0200, Abel Vesa wrote:
> Add support for the new i2c_pull property introduced for SM8550 setting
> a I2C specific pull mode on I2C able pins. Add the bit to the SM8550
> specific driver while at it.
>
> Co-developed-by: Neil Armstrong <neil.armstrong@xxxxxxxxxx>
> Signed-off-by: Neil Armstrong <neil.armstrong@xxxxxxxxxx>
> Signed-off-by: Abel Vesa <abel.vesa@xxxxxxxxxx>
> ---
>
> The v2 of this specific patch is here:
> https://lore.kernel.org/all/20221123152001.694546-4-abel.vesa@xxxxxxxxxx/
>
> Changes since v2:
> * This time, this patch is sent separate w.r.t. SM8550 pinctrl driver
> * The qcom,i2c-pull is dropped, bias-pull-up with value is used instead
> * Default value for i2c pull up is 2.2kOhms and since SM8550 is the
> first one to use it, we hard code it for now
> * changed the authorship as the implementation looks entirely different now
>
> drivers/pinctrl/qcom/pinctrl-msm.c | 7 +++++++
> drivers/pinctrl/qcom/pinctrl-msm.h | 1 +
> drivers/pinctrl/qcom/pinctrl-sm8550.c | 1 +
> 3 files changed, 9 insertions(+)
>
> diff --git a/drivers/pinctrl/qcom/pinctrl-msm.c b/drivers/pinctrl/qcom/pinctrl-msm.c
> index 5142c363480a..510c964dd0f5 100644
> --- a/drivers/pinctrl/qcom/pinctrl-msm.c
> +++ b/drivers/pinctrl/qcom/pinctrl-msm.c
> @@ -310,6 +310,8 @@ static int msm_config_reg(struct msm_pinctrl *pctrl,
> case PIN_CONFIG_BIAS_PULL_UP:
> *bit = g->pull_bit;
> *mask = 3;
> + if (g->i2c_pull_bit)
> + *mask |= BIT(g->i2c_pull_bit) >> *bit;
> break;
> case PIN_CONFIG_DRIVE_OPEN_DRAIN:
> *bit = g->od_bit;
> @@ -336,6 +338,7 @@ static int msm_config_reg(struct msm_pinctrl *pctrl,
> #define MSM_KEEPER 2
> #define MSM_PULL_UP_NO_KEEPER 2
> #define MSM_PULL_UP 3
> +#define MSM_I2C_STRONG_PULL_UP 2200
>
> static unsigned msm_regval_to_drive(u32 val)
> {
> @@ -387,6 +390,8 @@ static int msm_config_group_get(struct pinctrl_dev *pctldev,
> case PIN_CONFIG_BIAS_PULL_UP:
> if (pctrl->soc->pull_no_keeper)
> arg = arg == MSM_PULL_UP_NO_KEEPER;
> + else if (arg & BIT(g->i2c_pull_bit))
> + arg = MSM_I2C_STRONG_PULL_UP;
> else
> arg = arg == MSM_PULL_UP;
> if (!arg)
> @@ -467,6 +472,8 @@ static int msm_config_group_set(struct pinctrl_dev *pctldev,
> case PIN_CONFIG_BIAS_PULL_UP:
> if (pctrl->soc->pull_no_keeper)
> arg = MSM_PULL_UP_NO_KEEPER;
> + else if (g->i2c_pull_bit && arg > 1)

I would prefer that we do arg == MSM_I2C_STRONG_PULL_UP, primarily to
improve the symmetry with the getter, but also just to ensure that
someone accidentally writing bias-pull-up = <1>; gets less of a
surprise.

Looks good otherwise.

Thanks,
Bjorn

> + arg = BIT(g->i2c_pull_bit) | MSM_PULL_UP;
> else
> arg = MSM_PULL_UP;
> break;
> diff --git a/drivers/pinctrl/qcom/pinctrl-msm.h b/drivers/pinctrl/qcom/pinctrl-msm.h
> index 05a1209bf9ae..985eceda2517 100644
> --- a/drivers/pinctrl/qcom/pinctrl-msm.h
> +++ b/drivers/pinctrl/qcom/pinctrl-msm.h
> @@ -80,6 +80,7 @@ struct msm_pingroup {
>
> unsigned pull_bit:5;
> unsigned drv_bit:5;
> + unsigned i2c_pull_bit:5;
>
> unsigned od_bit:5;
> unsigned egpio_enable:5;
> diff --git a/drivers/pinctrl/qcom/pinctrl-sm8550.c b/drivers/pinctrl/qcom/pinctrl-sm8550.c
> index 0b7db7d4054a..c9d038098f2c 100644
> --- a/drivers/pinctrl/qcom/pinctrl-sm8550.c
> +++ b/drivers/pinctrl/qcom/pinctrl-sm8550.c
> @@ -47,6 +47,7 @@
> .mux_bit = 2, \
> .pull_bit = 0, \
> .drv_bit = 6, \
> + .i2c_pull_bit = 13, \
> .egpio_enable = 12, \
> .egpio_present = 11, \
> .oe_bit = 9, \
> --
> 2.34.1
>