Re: [PATCH v4 06/14] software node: get rid of property_set_pointer()

From: Andy Shevchenko
Date: Wed Sep 11 2019 - 05:21:47 EST


On Wed, Sep 11, 2019 at 01:37:25AM -0700, Dmitry Torokhov wrote:
> On Wed, Sep 11, 2019 at 11:29:10AM +0300, Andy Shevchenko wrote:
> > On Wed, Sep 11, 2019 at 8:15 AM Dmitry Torokhov
> > <dmitry.torokhov@xxxxxxxxx> wrote:
> > >
> > > Instead of explicitly setting values of integer types when copying
> > > property entries lets just copy entire value union when processing
> > > non-array values.
> > >
> > > When handling array values assign the pointer there using the newly
> > > introduced "raw" pointer union member. This allows us to remove
> > > property_set_pointer().
> > >
> > > In property_get_pointer() we do not need to handle each data type
> > > separately, we can simply return either the raw pointer or pointer to
> > > values union.
> >
> > Same as before, typechecking is good thing to have for my point of view.
> > Others may have different opinions.
>
> OK, I'll just point out that typechecking is a red herring here as
> everything was and still is accessed through void pointers, and we
> trusted the type set on property. Users of static properties should use
> PROPERTY_ENTRY_XXX() for initialization and do not poke into struct
> property_entry directly.
>
> I suppose it is up to Rafael to decide here.

Yes, and perhaps Mika as they were the main authors of the idea and
implementation.

--
With Best Regards,
Andy Shevchenko