Hi Mark, again
> > > + struct clk *clk;
> > > + u32 val;
> > > +
> > > + /*
> > > + * Parse dai->sysclk come from "clocks = <&xxx>"
> > > + * (if system has common clock)
> > > + * or "system-clock-frequency = <xxx>"
> > > + * or device's module clock.
> > > + */
> > > + clk = of_clk_get(port_np, 0);
> > > + if (!IS_ERR(clk)) {
> > > + simple_dai->sysclk = clk_get_rate(clk);
> > > + simple_dai->clk = clk;
> > > + } else if (!of_property_read_u32(port_np, "system-clock-frequency",
> > > &val)) {
> > > + simple_dai->sysclk = val;
> > > + } else {
> > > + clk = of_clk_get(endpoint_np, 0);
> > > + if (!IS_ERR(clk))
> > > + simple_dai->sysclk = clk_get_rate(clk);
> > > + }
> >
> > This looks like we're leaking the clocks - devm_ might help here
> > perhaps?
>
> Good catch.
> This came from original simple-card, but yes, we should use devm_*
> I will fix it on v3
Oops, of_clk_get() doesn't have devm_of_clk_get() ?
(and no of_clk_put() ... )
I will keep above as-is in v3. We can fix it incrementally (?)
