[PATCH v3 07/14] mfd: mt6370: Add Mediatek MT6370 support

[ChiaEn Wu]

Hi Lee,

Lee Jones <lee.jones at linaro.org> 於 2022年7月13日 週三 下午4:04寫道：
>
> On Wed, 13 Jul 2022, ChiaEn Wu wrote:
>
> > Hi Lee,
> >
> > Thanks for your reply.
> >
> > Lee Jones <lee.jones at linaro.org> 於 2022年7月12日 週二 晚上11:29寫道：
> > >
> > > On Thu, 23 Jun 2022, ChiaEn Wu wrote:
> > >
> > > > From: ChiYuan Huang <cy_huang at richtek.com>
> > > >
> > > > Add Mediatek MT6370 MFD support.
> > >
> > > No such thing as "MFD support".
> > >
> > > And you're not getting away with submitting a 370 line patch with a 5
> > > word change log either. :)
> > >
> > > Please at least tell us what the device is and what it's used for.
> >
> > I sincerely apologize.
> > We will add more descriptions of the MT6370 feature in the v5 patch.
> >
> > >
> > > > Signed-off-by: ChiYuan Huang <cy_huang at richtek.com>
> > > > ---
> > > >
> > > > v3
> > > > - Refine Kconfig help text
> > > > - Refine error message of unknown vendor ID in
> > > >   mt6370_check_vendor_info()
> > > > - Refine return value handling of mt6370_regmap_read()
> > > > - Refine all probe error by using dev_err_probe()
> > > > - Refine "bank_idx" and "bank_addr" in mt6370_regmap_read() and
> > > >   mt6370_regmap_write()
> > > > - Add "#define VENID*" and drop the comments in
> > > >   mt6370_check_vendor_info()
> > > > - Drop "MFD" in MODULE_DESCRIPTION()
> > > > ---
> > > >  drivers/mfd/Kconfig  |  13 ++
> > > >  drivers/mfd/Makefile |   1 +
> > > >  drivers/mfd/mt6370.c | 358 +++++++++++++++++++++++++++++++++++++++++++++++++++
> > > >  3 files changed, 372 insertions(+)
> > > >  create mode 100644 drivers/mfd/mt6370.c
>
> [...]
>
> > > > +static const struct mfd_cell mt6370_devices[] = {
> > > > +     MFD_CELL_OF("adc", NULL, NULL, 0, 0, "mediatek,mt6370-adc"),
> > > > +     MFD_CELL_OF("charger", NULL, NULL, 0, 0, "mediatek,mt6370-charger"),
> > > > +     MFD_CELL_OF("backlight", NULL, NULL, 0, 0, "mediatek,mt6370-backlight"),
> > > > +     MFD_CELL_OF("flashlight", NULL, NULL, 0, 0, "mediatek,mt6370-flashlight"),
> > > > +     MFD_CELL_OF("indicator", NULL, NULL, 0, 0, "mediatek,mt6370-indicator"),
> > > > +     MFD_CELL_OF("tcpc", NULL, NULL, 0, 0, "mediatek,mt6370-tcpc"),
> > > > +     MFD_CELL_RES("regulator", mt6370_regulator_irqs)
> > >
> > > The first parameters here should be prepended with something, perhaps
> > > "mt6370_"?
>
> > OK, we will add the prefix in the next patch.

Sorry, I forgot to ask a question in the last mail.
I wonder if using "mt6370-xxx" (dash) is better than using "mt6370_"
(underline) ??
Thanks.

> [...]
>
> > > > +static int mt6370_probe(struct i2c_client *i2c)
> > > > +{
> > > > +     struct mt6370_info *info;
> > > > +     struct i2c_client *usbc_i2c;
> > > > +     int ret;
> > > > +
> > > > +     info = devm_kzalloc(&i2c->dev, sizeof(*info), GFP_KERNEL);
> > > > +     if (!info)
> > > > +             return -ENOMEM;
> > > > +
> > > > +     info->dev = &i2c->dev;
> > > > +
> > > > +     usbc_i2c = devm_i2c_new_dummy_device(&i2c->dev, i2c->adapter,
> > > > +                                          MT6370_USBC_I2CADDR);
> > > > +     if (IS_ERR(usbc_i2c))
> > > > +             return dev_err_probe(&i2c->dev, PTR_ERR(usbc_i2c),
> > > > +                                  "Failed to register USBC I2C client\n");
> > > > +
> > > > +     /* Assign I2C client for PMU and TypeC */
> > > > +     info->i2c[MT6370_PMU_I2C] = i2c;
> > > > +     info->i2c[MT6370_USBC_I2C] = usbc_i2c;
> > > > +
> > > > +     info->regmap = devm_regmap_init(&i2c->dev, &mt6370_regmap_bus, info,
> > > > +                                     &mt6370_regmap_config);
> > >
> > > Apart from in mt6370_check_vendor_info() where is this actually used?
> >
> > Well... from my understanding, we use this MFD driver to make other
> > drivers of MT6370 (e.g. charger, ADC, led...) use the same regmap
> > settings.
> > Thus, this regmap is not only used in mt6370_check_vendor_info().
>
> Well for that to happen you need to store the data somewhere for the
> child devices to fetch from.  I don't see that happening in this
> patch?  What did I miss?

hmmm... I got your point... I will let regmap be a local var in
probe() in the next patch.
Thank you so much!

>
> --
> Lee Jones [李琼斯]
> Principal Technical Lead - Developer Services
> Linaro.org │ Open source software for Arm SoCs
> Follow Linaro: Facebook | Twitter | Blog

Best regards,
ChiaEn Wu