[RESEND. PATCH v1] pinctrl: mediatek: paris: Revert "Rework support for PIN_CONFIG_{INPUT,OUTPUT}_ENABLE"

Macpaul Lin macpaul.lin at mediatek.com
Thu Oct 17 06:07:51 PDT 2024



On 10/17/24 17:14, Bo Ye wrote:
> [This reverts commit c5d3b64c568a344e998830e0e94a7c04e372f89b.]

First, this should be "PATCH v2" since you've already modified commit 
message.

> For MTK HW,
> 1. to enable GPIO input direction: set DIR=0, IES=1
> 2. to enable GPIO output direction: set DIR=1, and set DO=1 to output high, set DO=0 to out low
> 
> The PIN_CONFIG_INPUT/PIN_CONFIG_OUTPUT/PIN_CONFIG_INPUT_ENABLE/PIN_CONFIG_OUTPUT_ENABLE shall
> be implemented according to view of its purpose - set GPIO direction and output value (for
> output only) according to specific HW design.
> 
> However, the reverted patch implement according to author's own explanation of IES without
> understanding of MTK's HW. Such patch does not correctly set DIR/IES bit to control GPIO
> direction on MTK's HW.
> 
> Fixes: c5d3b64c568 ("pinctrl: mediatek: paris: Rework support for PIN_CONFIG_{INPUT,OUTPUT}_ENABLE")
> Signed-off-by: Light Hsieh <light.hsieh at mediatek.com>
> Signed-off-by: Evan Cao <ot_evan.cao at mediatek.com>
> Signed-off-by: Bo Ye <bo.ye at mediatek.com>
> ---
>   drivers/pinctrl/mediatek/pinctrl-paris.c | 38 +++++++++++++++++-------
>   1 file changed, 27 insertions(+), 11 deletions(-)

Please add change logs in this section.
For example:

Changes for v2:
  - Update "Fixes: " tag in commit message.

> diff --git a/drivers/pinctrl/mediatek/pinctrl-paris.c b/drivers/pinctrl/mediatek/pinctrl-paris.c
> index 87e958d827bf..a8af62e6f8ca 100644
> --- a/drivers/pinctrl/mediatek/pinctrl-paris.c
> +++ b/drivers/pinctrl/mediatek/pinctrl-paris.c
> @@ -165,21 +165,20 @@ static int mtk_pinconf_get(struct pinctrl_dev *pctldev,

[snip]

Please reply all the questions address by the reviewers.
For example:
[1] 
https://lore.kernel.org/lkml/433295fe-8d34-af8b-f6bf-be1953b6e479@mediatek.com/T/#m168c6138c374009990025b0407f617ba10dede8d

Otherwise this patch will hard to be reviewed.
Any open-minded discussion contribute to the improvement of Linux code 
and the community.

Thanks
Macpaul Lin



More information about the linux-arm-kernel mailing list