Re: [PATCH 2/2] clk: core: link consumer with clock driver

[Stephen Boyd <sboyd@kernel.org>]

Quoting Miquel Raynal (2018-11-30 02:20:52)
> Hi Stephen,
> 
> Stephen Boyd <sboyd@kernel.org> wrote on Fri, 30 Nov 2018 01:26:20
> -0800:
> 
> > Quoting Miquel Raynal (2018-11-23 01:11:32)
> > > Would you agree with me adding dummy functions in the #else section
> > > like:
> > > 
> > > static inline void __clk_device_link(struct device *consumer, struct clk *clk)
> > > {
> > >        return;
> > > }
> > > 
> > > static inline void __clk_device_unlink(struct clk *clk)
> > > {
> > >        return;
> > > }
> > > 
> > > Do you want me to also declare these functions in the #if section
> > > (with the external keyword) to balance the above declarations?  
> > 
> > Why can't we do the linking in __clk_get() and __clk_put()?
> > 
> 
> Because we need the caller's 'struct device' to make the link and
> this is not available in __clk_get(). I tried to ad it as parameter but
> I don't think it is possible to retrieve a 'struct device' from the
> device name. The functions where this is problematic are:
> * clk.c:__of_clk_get_from_provider()
> * clkdev.c:clk_get_sys()
> 
> By the way in my new version I called the helpers:
> * clk_{link,unlink}_hierarchy()
> * clk_{link,unlink}_consumer()
> 
> I will send a new version with these helpers, but if you have anything
> in mind to help me achieve the above request, I will welcome the idea.
> 

We can do the linking in __clk_get() and __clk_put() if we poke into the
struct clk -> struct clk_core and bury the struct device into each
clk_core structure.