Re: [PATCH 2/2] soc: mediatek: devapc: Add support for MT8186

[AngeloGioacchino Del Regno]

Il 29/04/22 08:54, Rex-BC Chen ha scritto:
> Add devapc data and compatible to support MT8186 devapc.
>
> Signed-off-by: Runyang Chen <runyang.chen@xxxxxxxxxxxx>
> Signed-off-by: Rex-BC Chen <rex-bc.chen@xxxxxxxxxxxx>
> ---
>   drivers/soc/mediatek/mtk-devapc.c | 15 +++++++++++++++
>   1 file changed, 15 insertions(+)
>
> diff --git a/drivers/soc/mediatek/mtk-devapc.c b/drivers/soc/mediatek/mtk-devapc.c
> index 7c65ad3d1f8a..762e0e4d967d 100644
> --- a/drivers/soc/mediatek/mtk-devapc.c
> +++ b/drivers/soc/mediatek/mtk-devapc.c
> @@ -227,10 +227,25 @@ static const struct mtk_devapc_data devapc_mt6779 = {
>   	.vio_shift_con_offset = 0xF20,
>   };
>   
> +static const struct mtk_devapc_data devapc_mt8186 = {
> +	.vio_idx_num = 519,
> +	.vio_mask_offset = 0x0,
> +	.vio_sta_offset = 0x400,
> +	.vio_dbg0_offset = 0x900,
> +	.vio_dbg1_offset = 0x904,
> +	.apc_con_offset = 0xF00,
> +	.vio_shift_sta_offset = 0xF10,
> +	.vio_shift_sel_offset = 0xF14,
> +	.vio_shift_con_offset = 0xF20,
> +};

Hello Rex,
since the register layout is common across a lot of MediaTek SoCs and basically
only the vio_idx_num wildly varies between them, can you separate this parameter
from the mtk_devapc_data struct?

I propose the following:

struct mtk_devapc_regs {
	/* reg offset */
	u32 vio_mask_offset;
	u32 ... etc
};

struct mtk_devapc_data {
	u32 vio_idx_num;
	const struct mtk_devapc_regs *regs;
};

static const struct mtk_devapc_regs devapc_regs_mt6779 = {
	.vio_mask_offset = 0x0,
	.vio...etc
};

static const struct mtk_devapc_data devapc_mt6779 = {
	.vio_idx_num = 511,
	.regs = devapc_regs_mt6779,
};

static const struct mtk_devapc_data devapc_mt8186 = {
	.vio_idx_num = 519,
	.regs = devapc_regs_mt6779,
};

Regards,
Angelo