Re: [PATCH 4.19 29/74] clk: qcom: regmap-mux: fix parent clock lookup

[Dmitry Baryshkov]

On 15/12/2021 12:16, Pavel Machek wrote:
> Hi!
>
> > The function mux_get_parent() uses qcom_find_src_index() to find the
> > parent clock index, which is incorrect: qcom_find_src_index() uses src
> > enum for the lookup, while mux_get_parent() should use cfg field (which
> > corresponds to the register value). Add qcom_find_cfg_index() function
> > doing this kind of lookup and use it for mux parent lookup.
>
> This appears to have problems with error handling.
>
> > +++ b/drivers/clk/qcom/clk-regmap-mux.c
> > @@ -36,7 +36,7 @@ static u8 mux_get_parent(struct clk_hw *
> >   	val &= mask;
> >   
> >   	if (mux->parent_map)
> > -		return qcom_find_src_index(hw, mux->parent_map, val);
> > +		return qcom_find_cfg_index(hw, mux->parent_map, val);
> >   
> >   	return val;
> >   }
>
> So this returns u8.
>
> > +int qcom_find_cfg_index(struct clk_hw *hw, const struct parent_map *map, u8 cfg)
> > +{
> > +	int i, num_parents = clk_hw_get_num_parents(hw);
> > +
> > +	for (i = 0; i < num_parents; i++)
> > +		if (cfg == map[i].cfg)
> > +			return i;
> > +
> > +	return -ENOENT;
> > +}
>
> In case of error, -ENOENT will be cast to u8 in caller. I don't
> believe that is correct.

Unfortunately there is no way to return proper error code from 
clk_ops->get_parent() callback. However returning -ENOENT would 
translate to 254. Then clk_core_get_parent_by_index() would determine 
that there is no such parent and return NULL. A call to clk_set_parent 
would reparent the clock.

Returning some sensible default (e.g. 0) would be much worse, since then 
the clock subsystem would assume that the clock has correct parent. A 
call to clk_set_parent would always result in ops->set_parent() call, 
reparenting the clock correctly.

Most probably it would be correct to make ops->get_parent() return int 
instead of u8 (either an index or an -ERROR). However this was out of 
scope for this patch.

--
With best wishes
Dmitry