Re: [PATCH 01/18] phy: mediatek: add a new helper to update bitfield

[AngeloGioacchino Del Regno <angelogioacchino.delregno@collabora.com>]

Il 22/09/22 04:36, Chunfeng Yun ha scritto:
> On Wed, 2022-09-21 at 10:15 +0200, AngeloGioacchino Del Regno wrote:
>> Il 20/09/22 11:00, Chunfeng Yun ha scritto:
>>> Due to FIELD_PREP() macro can be used to prepare a bitfield value,
>>> local ones can be remove; add the new helper to make bitfield
>>> update
>>> easier.
>>>
>>> Signed-off-by: Chunfeng Yun <chunfeng.yun@mediatek.com>
>>> ---
>>>    drivers/phy/mediatek/phy-mtk-io.h | 7 +++++++
>>>    1 file changed, 7 insertions(+)
>>>
>>> diff --git a/drivers/phy/mediatek/phy-mtk-io.h
>>> b/drivers/phy/mediatek/phy-mtk-io.h
>>> index 500fcdab165d..a723d4afc9b4 100644
>>> --- a/drivers/phy/mediatek/phy-mtk-io.h
>>> +++ b/drivers/phy/mediatek/phy-mtk-io.h
>>> @@ -8,6 +8,7 @@
>>>    #ifndef __PHY_MTK_H__
>>>    #define __PHY_MTK_H__
>>>    
>>> +#include <linux/bitfield.h>
>>>    #include <linux/io.h>
>>>    
>>>    static inline void mtk_phy_clear_bits(void __iomem *reg, u32
>>> bits)
>>> @@ -35,4 +36,10 @@ static inline void mtk_phy_update_bits(void
>>> __iomem *reg, u32 mask, u32 val)
>>>    	writel(tmp, reg);
>>>    }
>>>    
>>> +/* field @mask should be constant and continuous */
>>
>> "Field @mask shall be [...]"
>>                ^^^^^
> Ok, will modify it
> 
>>
>>> +static inline void mtk_phy_update_field(void __iomem *reg, u32
>>> mask, u32 val)
>>
>> ...so, (void __iomem *reg, const u32 mask, u32 val)
> Maybe no need const, @mask will be checked it in compile time when
> use FIELD_PREP(), means @mask is a constant value, but not a variable.
> 

Adding const is not *required*, but `mask` is, effectively, a constant, hence
it fully makes sense to add const.

> Thanks
> 
>>
>>> +{
>>> +	mtk_phy_update_bits(reg, mask, FIELD_PREP(mask, val));
>>> +}
>>> +
>>>    #endif
>>
>>
>