On 02/03/2021 08:16, Gene Chen wrote:
> Lee Jones <lee.jones@xxxxxxxxxx> 於 2021年1月15日 週五 下午3:32寫道：
>>
>> On Fri, 15 Jan 2021, Gene Chen wrote:
>>
>>> Matthias Brugger <matthias.bgg@xxxxxxxxx> 於 2021年1月12日 週二 下午8:32寫道：
>>>>
>>>>
>>>>
>>>> On 12/11/2020 11:39, Gene Chen wrote:
>>>>> From: Gene Chen <gene_chen@xxxxxxxxxxx>
>>>>>
>>>>> Combine mt6360 pmic/ldo resources into mt6360 regulator resources
>>>>> to simplify the similar resources object.
>>>>>
>>>>> Signed-off-by: Gene Chen <gene_chen@xxxxxxxxxxx>
>>>>> Acked-for-MFD-by: Lee Jones <lee.jones@xxxxxxxxxx>
>>>>> ---
>>>>> drivers/mfd/mt6360-core.c | 11 +++--------
>>>>> 1 file changed, 3 insertions(+), 8 deletions(-)
>>>>>
>>>>> diff --git a/drivers/mfd/mt6360-core.c b/drivers/mfd/mt6360-core.c
>>>>> index 692e47b..5119e51 100644
>>>>> --- a/drivers/mfd/mt6360-core.c
>>>>> +++ b/drivers/mfd/mt6360-core.c
>>>>> @@ -265,7 +265,7 @@ static const struct resource mt6360_led_resources[] = {
>>>>> DEFINE_RES_IRQ_NAMED(MT6360_FLED1_STRB_TO_EVT, "fled1_strb_to_evt"),
>>>>> };
>>>>>
>>>>> -static const struct resource mt6360_pmic_resources[] = {
>>>>> +static const struct resource mt6360_regulator_resources[] = {
>>>>> DEFINE_RES_IRQ_NAMED(MT6360_BUCK1_PGB_EVT, "buck1_pgb_evt"),
>>>>> DEFINE_RES_IRQ_NAMED(MT6360_BUCK1_OC_EVT, "buck1_oc_evt"),
>>>>> DEFINE_RES_IRQ_NAMED(MT6360_BUCK1_OV_EVT, "buck1_ov_evt"),
>>>>> @@ -278,9 +278,6 @@ static const struct resource mt6360_pmic_resources[] = {
>>>>> DEFINE_RES_IRQ_NAMED(MT6360_LDO7_OC_EVT, "ldo7_oc_evt"),
>>>>> DEFINE_RES_IRQ_NAMED(MT6360_LDO6_PGB_EVT, "ldo6_pgb_evt"),
>>>>> DEFINE_RES_IRQ_NAMED(MT6360_LDO7_PGB_EVT, "ldo7_pgb_evt"),
>>>>> -};
>>>>> -
>>>>> -static const struct resource mt6360_ldo_resources[] = {
>>>>> DEFINE_RES_IRQ_NAMED(MT6360_LDO1_OC_EVT, "ldo1_oc_evt"),
>>>>> DEFINE_RES_IRQ_NAMED(MT6360_LDO2_OC_EVT, "ldo2_oc_evt"),
>>>>> DEFINE_RES_IRQ_NAMED(MT6360_LDO3_OC_EVT, "ldo3_oc_evt"),
>>>>> @@ -298,10 +295,8 @@ static const struct mfd_cell mt6360_devs[] = {
>>>>> NULL, 0, 0, "mediatek,mt6360-chg"),
>>>>> OF_MFD_CELL("mt6360-led", mt6360_led_resources,
>>>>> NULL, 0, 0, "mediatek,mt6360-led"),
>>>>> - OF_MFD_CELL("mt6360-pmic", mt6360_pmic_resources,
>>>>> - NULL, 0, 0, "mediatek,mt6360-pmic"),
>>>>> - OF_MFD_CELL("mt6360-ldo", mt6360_ldo_resources,
>>>>> - NULL, 0, 0, "mediatek,mt6360-ldo"),
>>>>> + OF_MFD_CELL("mt6360-regulator", mt6360_regulator_resources,
>>>>> + NULL, 0, 0, "mediatek,mt6360-regulator"),
>>>>
>>>> As discussed with the MFD maintainer [1], the regulator (and probably all cells)
>>>> shouldn't have a DT binding.
>>>>
>>>> So please send a new version which fixes that.
>>>>
>>>> Regards,
>>>> Matthias
>>>>
>>>> [1] https://lore.kernel.org/linux-mediatek/20210111164118.GE4728@xxxxxxxxxxxxx/
>>
>> I don't think Mark is correct here.
>>
>> We usually do implement compatible strings for sub-devices and they do
>> tend to have their own device nodes.
>>
>> It's a very long time ago since I coded this up myself, but from
>> memory, you can't have 2 devices share a compatible string.
>>
>
> Either Mark or Lee suggestion is work.
> Is there a conclusion that we can apply it?
> If MFD is already supported of_compatible, we prefer use of_compatible mapping.
>
For regulator on PMIC you should add the regulator to the pmic node directly.
Please see the series from Hsin-Hsiung for referece:
https://lore.kernel.org/linux-mediatek/1615829757-3223-6-git-send-email-hsin-hsiung.wang@xxxxxxxxxxxx/T/#u
Regards,
Matthias
>>> Should I use parent's device to find sub-devices of_node if without
>>> compatible name?
>>> I trace the function mfd_add_device,
>>>
>>> if (IS_ENABLED(CONFIG_OF) && parent->of_node && cell->of_compatible) {
>>> .....
>>> ret = mfd_match_of_node_to_dev(pdev, np, cell);
>>> .....
>>> }
>>>
>>> which is binding mfd sub-device with compatible. Does it be removed in
>>> the feature?
>>>
>>>>> OF_MFD_CELL("mt6360-tcpc", NULL,
>>>>> NULL, 0, 0, "mediatek,mt6360-tcpc"),
>>>>> };
>>>>>
>>
>> --
>> Lee Jones [李琼斯]
>> Senior Technical Lead - Developer Services
>> Linaro.org │ Open source software for Arm SoCs
>> Follow Linaro: Facebook | Twitter | Blog
