Re: [PATCH 05/10] mfd: rt5033: Apply preparatory changes before adding rt5033-charger driver

[Lee Jones <lee@kernel.org>]

On Tue, 28 Feb 2023, Jakob Hauser wrote:

> Order the register blocks to have the masks in descending manner.
> 
> Add new defines for constant voltage shift (RT5033_CHGCTRL2_CV_SHIFT),
> MIVR mask (RT5033_CHGCTRL4_MIVR_MASK), pre-charge current shift
> (RT5033_CHGCTRL4_IPREC_SHIFT), internal timer disable
> (RT5033_INT_TIMER_DISABLE), termination disable (RT5033_TE_DISABLE),
> CFO disable (RT5033_CFO_DISABLE), UUG disable (RT5033_CHARGER_UUG_DISABLE).
> 
> The fast charge timer type needs to be written on mask 0x38
> (RT5033_CHGCTRL3_TIMER_MASK). To avoid a bit shift on application, change the
> values of the timer types to fit the mask. Added the timout duration as a
> comment. And the timer between TIMER8 and TIMER12 is most likely TIMER10, see
> e.g. RT5036 [1] page 28 bottom.
> 
> Add value options for MIVR (Minimum Input Voltage Regulation).
> 
> Move RT5033_TE_ENABLE_MASK to the block "RT5033 CHGCTRL1 register", in order
> to have the masks of the register collected there. To fit the naming scheme,
> rename it to RT5033_CHGCTRL1_TE_EN_MASK.
> 
> Move RT5033_CHG_MAX_CURRENT to the block "RT5033 charger fast-charge current".
> 
> Add new defines RT5033_CV_MAX_VOLTAGE and RT5033_CHG_MAX_PRE_CURRENT to the
> blocks "RT5033 charger constant charge voltage" and "RT5033 charger pre-charge
> current limits".
> 
> In include/linux/mfd/rt5033.h, turn power_supply "psy" into a pointer in order
> to use it in devm_power_supply_register().

Are there no present users to account for?

> [1] https://media.digikey.com/pdf/Data%20Sheets/Richtek%20PDF/RT5036%20%20Preliminary.pdf
> 
> Signed-off-by: Jakob Hauser <jahau@rocketmail.com>
> ---
>  include/linux/mfd/rt5033-private.h | 53 ++++++++++++++++++++----------
>  include/linux/mfd/rt5033.h         |  2 +-
>  2 files changed, 36 insertions(+),` 19 deletions(-)

-- 
Lee Jones [李琼斯]