Re: [PATCH v3 1/2] driver: reset: spacemit-p1: add driver for poweroff/reboot

[Aurelien Jarno <aurelien@aurel32.net>]

On 2025-10-22 08:48, Yixun Lan wrote:
> Hi Aurelien,
> 
> On 22:11 Tue 21 Oct     , Aurelien Jarno wrote:
> > This driver implements poweroff/reboot support for the SpacemiT P1 PMIC
> > chip, which is commonly paired with the SpacemiT K1 SoC.
> > 
> > The SpacemiT P1 support is implemented as a MFD driver, so the access is
> > done directly through the regmap interface. Reboot or poweroff is
> > triggered by setting a specific bit in a control register, which is
> > automatically cleared by the hardware afterwards.
> > 
> > Signed-off-by: Aurelien Jarno <aurelien@aurel32.net>
> > Reviewed-by: Yixun Lan <dlan@gentoo.org>
> > ---
> > v3:
> >  - Allow building as a module
> >  - Remove outdated Acked-by and Tested-by
> >  - Collect Reviewed-by
> > 
> >  drivers/power/reset/Kconfig              |  9 +++
> >  drivers/power/reset/Makefile             |  1 +
> >  drivers/power/reset/spacemit-p1-reboot.c | 88 ++++++++++++++++++++++++
> >  3 files changed, 98 insertions(+)
> >  create mode 100644 drivers/power/reset/spacemit-p1-reboot.c
> > 
> > diff --git a/drivers/power/reset/Kconfig b/drivers/power/reset/Kconfig
> > index 8248895ca9038..6577d73edbda4 100644
> > --- a/drivers/power/reset/Kconfig
> > +++ b/drivers/power/reset/Kconfig
> > @@ -283,6 +283,15 @@ config POWER_RESET_KEYSTONE
> >  	help
> >  	  Reboot support for the KEYSTONE SoCs.
> >  
> > +config POWER_RESET_SPACEMIT_P1
> > +	tristate "SpacemiT P1 poweroff and reset driver"
> > +	depends on ARCH_SPACEMIT || COMPILE_TEST
> ..
> > +	select MFD_SPACEMIT_P1
> I'd suggest to use "depends on" instead of "select", the reason is that
> using "select" will sometimes ignore the dependency, considering
> the reset driver here is tightly coupled with P1, so I think it's 
> reasonable to switch to use "depends on", also refer below link
> 
> https://lxr.linux.no/#linux+v6.7.1/Documentation/kbuild/kconfig-language.rst#L144
> 
>         select should be used with care. select will force
>         a symbol to a value without visiting the dependencies.
>         By abusing select you are able to select a symbol FOO even
>         if FOO depends on BAR that is not set.
>         In general use select only for non-visible symbols
>         (no prompts anywhere) and for symbols with no dependencies.
>         That will limit the usefulness but on the other hand avoid
>         the illegal configurations all over.

Thanks for the pointer, I'll fix that in the next version. I used 
REGULATOR_SPACEMIT_P1 and RTC_DRV_SPACEMIT_P1 as examples, they'll also 
need to be fixed.

Note also that without the select, a default value has to be added to 
MFD_SPACEMIT_P1, otherwise this makes the default values on the 
regulator, rtc and reboot drivers useless.

-- 
Aurelien Jarno                          GPG: 4096R/1DDD8C9B
aurelien@aurel32.net                     http://aurel32.net