Re: [PATCH 2/7] pinctrl: mediatek: paris: Fix PIN_CONFIG_BIAS_DISABLE readback

[Guodong Liu <guodong.liu@xxxxxxxxxxxx>]

-----Original Message-----
From: Chen-Yu Tsai <wenst@xxxxxxxxxxxx>
To: Guodong Liu <guodong.liu@xxxxxxxxxxxx>
Cc: Sean Wang <sean.wang@xxxxxxxxxx>, Linus Walleij <
linus.walleij@xxxxxxxxxx>, Matthias Brugger <matthias.bgg@xxxxxxxxx>, 
linux-mediatek@xxxxxxxxxxxxxxxxxxx, linux-gpio@xxxxxxxxxxxxxxx, 
linux-arm-kernel@xxxxxxxxxxxxxxxxxxx, linux-kernel@xxxxxxxxxxxxxxx,
Zhiyong Tao <zhiyong.tao@xxxxxxxxxxxx>, Hui Liu <hui.liu@xxxxxxxxxxxx>,
Light Hsieh <light.hsieh@xxxxxxxxxxxx>
Subject: Re: [PATCH 2/7] pinctrl: mediatek: paris: Fix
PIN_CONFIG_BIAS_DISABLE readback
Date: Wed, 19 Jan 2022 13:57:18 +0800

On Wed, Jan 19, 2022 at 9:42 AM Guodong Liu <guodong.liu@xxxxxxxxxxxx>
wrote:
> 
> -----Original Message-----
> From: Chen-Yu Tsai <wenst@xxxxxxxxxxxx>
> To: Sean Wang <sean.wang@xxxxxxxxxx>, Linus Walleij <
> linus.walleij@xxxxxxxxxx>, Matthias Brugger <matthias.bgg@xxxxxxxxx>
> Cc: Chen-Yu Tsai <wenst@xxxxxxxxxxxx>,
> linux-mediatek@xxxxxxxxxxxxxxxxxxx, linux-gpio@xxxxxxxxxxxxxxx,
> linux-arm-kernel@xxxxxxxxxxxxxxxxxxx, linux-kernel@xxxxxxxxxxxxxxx,
> Zhiyong Tao <zhiyong.tao@xxxxxxxxxxxx>, Guodong Liu <
> guodong.liu@xxxxxxxxxxxx>
> Subject: [PATCH 2/7] pinctrl: mediatek: paris: Fix
> PIN_CONFIG_BIAS_DISABLE readback
> Date: Tue, 11 Jan 2022 19:22:39 +0800
> 
> When reading back pin bias settings, if the pin is not in a
> bias-disabled state, the function should return -EINVAL.
> 
> Fix this in the mediatek-paris pinctrl library so that the read back
> state is not littered with bogus a "input bias disabled" combined
> with
> "pull up" or "pull down" states.
> 
> Fixes: 805250982bb5 ("pinctrl: mediatek: add pinctrl-paris that
> implements the vendor dt-bindings")
> Signed-off-by: Chen-Yu Tsai <wenst@xxxxxxxxxxxx>
> ---
>  drivers/pinctrl/mediatek/pinctrl-paris.c | 4 ++--
>  1 file changed, 2 insertions(+), 2 deletions(-)
> 
> diff --git a/drivers/pinctrl/mediatek/pinctrl-paris.c
> b/drivers/pinctrl/mediatek/pinctrl-paris.c
> index f9f9110f2107..1ca598ea7ba7 100644
> --- a/drivers/pinctrl/mediatek/pinctrl-paris.c
> +++ b/drivers/pinctrl/mediatek/pinctrl-paris.c
> @@ -97,8 +97,8 @@ static int mtk_pinconf_get(struct pinctrl_dev
> *pctldev,
>                         if (err)
>                                 goto out;
>                         if (param == PIN_CONFIG_BIAS_DISABLE) {
> -                               if (ret == MTK_PUPD_SET_R1R0_00)
> -                                       ret = MTK_DISABLE;
> +                               if (ret != MTK_PUPD_SET_R1R0_00)
> +                                       err = -EINVAL;
> Hi Chen-Yu
> 
> When the API "hw->soc->bias_get_combo(hw, desc, &pullup, &ret)" is
> called,
> The ret vaule of ret may be MTK_DISABLE or MTK_PUPD_SET_R1R0_00
> or  (pullen
> == 0),  All those cases are expected to be as "bias-disable".
> We advices to keep original code,
> +                               if (ret == MTK_PUPD_SET_R1R0_00)
> +                                       ret = MTK_DISABLE;
> +                               if (ret != MTK_DISABLE)
> +                                       err = -EINVAL;

IIUC you are suggesting to assign MTK_DISABLE to ret in the other two
cases,
and then check if ret == MTK_DISABLE.

Thanks for pointing that out.

ChenYu

> Thanks

Hi Chen-Yu

Yes, just for pins with config of MTK_PUPD_SET_R1R0_00 are required to
do additional assignment operations(ret = MTK_DISABLE;), in the other
two cases, the assignment operations of ret as MTK_DISABLE is obtained
by function call "hw->soc->bias_get_combo(hw, desc, &pullup, &ret)".

Thanks

>                         } else if (param == PIN_CONFIG_BIAS_PULL_UP)
> {
>                                 /* When desire to get pull-up value,
> return
>                                  *  error if current setting is pull-
> down
>