Re: [PATCH v13 3/8] clk: Use devm_ in the register fixed factor clock

[Stephen Boyd]

Quoting Niklas Cassel (2018-10-14 13:21:22)
> On Fri, Oct 12, 2018 at 03:08:17PM -0700, Stephen Boyd wrote:
> > Quoting Ricardo Salveti (2018-09-14 11:53:02)
> > > On Thu, Jun 14, 2018 at 6:55 PM <ilia.lin@xxxxxxxxx> wrote:
> > > >
> > > > From: Ilia Lin <ilialin@xxxxxxxxxxxxxx>
> > > >
> > > > Use devm_clk_hw_register instead of clk_hw_register
> > > > to simplify the usage of this API. This way drivers that call
> > > > the clk_hw_register_fixed_factor won't need to maintain
> > > > a data structure for further cleanup.
> > > >
> > > > Signed-off-by: Ilia Lin <ilialin@xxxxxxxxxxxxxx>
> > > > Tested-by: Amit Kucheria <amit.kucheria@xxxxxxxxxx>
> > > > ---
> > > >  drivers/clk/clk-fixed-factor.c | 2 +-
> > > >  1 file changed, 1 insertion(+), 1 deletion(-)
> > > >
> > > > diff --git a/drivers/clk/clk-fixed-factor.c b/drivers/clk/clk-fixed-factor.c
> > > > index a5d402de5584..8e39bda8e596 100644
> > > > --- a/drivers/clk/clk-fixed-factor.c
> > > > +++ b/drivers/clk/clk-fixed-factor.c
> > > > @@ -94,7 +94,7 @@ struct clk_hw *clk_hw_register_fixed_factor(struct device *dev,
> > > >         init.num_parents = 1;
> > > >
> > > >         hw = &fix->hw;
> > > > -       ret = clk_hw_register(dev, hw);
> > > > +       ret = devm_clk_hw_register(dev, hw);
> > > 
> > > Not sure what is the current state of this patch-set, but this change
> > > breaks drivers calling clk_hw_register_fixed_factor with a NULL dev
> > > (e.g. imx_clk_fixed_factor), as devm_clk_hw_register needs a valid dev
> > > for devres_add to work.
> > > 
> > 
> > Yep. Probably better to just have a driver register the clk_hw structure
> > itself with the clk framework vs. trying to get it right here in the
> > generic type registration function.
> > 
> 
> Hello Ilia, Stephen,
> 
> As you know this patch series is needed to get DVFS working on msm8996.
> 
> Ilia, are you planning on submitting a new patch set for this, or would
> you prefer if we submitted a new patch version for you?
> 
> Looking at this patch series, this appears to be the only outstanding
> review comment.

Sorry, you were too hopeful :/

> Stephen, do you have any review comments on this series?
> 

Yes, my nitpick should be easy to resolve and then it's just a few open
questions after that. Otherwise things seem fine and the CPU can run
really fast without bumping up the voltages?