Re: [PATCH v4 03/22] ASoC: sun4i-i2s: Change get_sr() and get_wss() to be more explicit

[Clément Péron]

Hi Maxime,

On Mon, 21 Sep 2020 at 15:55, Maxime Ripard <maxime@xxxxxxxxxx> wrote:
>
> On Mon, Sep 21, 2020 at 12:27:12PM +0200, Clément Péron wrote:
> > We are actually using a complex formula to just return a bunch of
> > simple values. Also this formula is wrong for sun4i.
>
> Just like the previous patch, this could use a bit more explanation,
> like why it's a good thing, or how it's wrong on sun4i

Okay I will comment a bit more.

>
> > Replace this with a simpler switch case.
> >
> > Also drop the i2s params not used.
> >
> > Signed-off-by: Clément Péron <peron.clem@xxxxxxxxx>
> > ---
> >  sound/soc/sunxi/sun4i-i2s.c | 69 +++++++++++++++++++++++--------------
> >  1 file changed, 44 insertions(+), 25 deletions(-)
> >
> > diff --git a/sound/soc/sunxi/sun4i-i2s.c b/sound/soc/sunxi/sun4i-i2s.c
> > index 0633b9fba3d7..11bbcbe24d6b 100644
> > --- a/sound/soc/sunxi/sun4i-i2s.c
> > +++ b/sound/soc/sunxi/sun4i-i2s.c
> > @@ -175,8 +175,8 @@ struct sun4i_i2s_quirks {
> >       unsigned int                    num_mclk_dividers;
> >
> >       unsigned long (*get_bclk_parent_rate)(const struct sun4i_i2s *);
> > -     s8      (*get_sr)(const struct sun4i_i2s *, int);
> > -     s8      (*get_wss)(const struct sun4i_i2s *, int);
> > +     int     (*get_sr)(unsigned int width);
> > +     int     (*get_wss)(unsigned int width);
> >       int     (*set_chan_cfg)(const struct sun4i_i2s *i2s,
> >                               unsigned int channels,  unsigned int slots,
> >                               unsigned int slot_width);
> > @@ -381,37 +381,56 @@ static int sun4i_i2s_set_clk_rate(struct snd_soc_dai *dai,
> >       return 0;
> >  }
> >
> > -static s8 sun4i_i2s_get_sr(const struct sun4i_i2s *i2s, int width)
> > +static int sun4i_i2s_get_sr(unsigned int width)
> >  {
> > -     if (width < 16 || width > 24)
> > -             return -EINVAL;
> > -
> > -     if (width % 4)
> > -             return -EINVAL;
> > +     switch (width) {
> > +     case 16:
> > +             return 0x0;
> > +     case 20:
> > +             return 0x1;
> > +     case 24:
> > +             return 0x2;
> > +     }
>
> Why do we need an hex number here?

This is a register value, so I thought it's usually written using
hexadecimal representation.

>
> Also, why is the return type change needed?

This function returns a ERROR defined in errno.h which actually could
be -133 but S8 only supports -128..127.

There is no real reason to have a S8 here and doesn't give any optimisation.

Clement

>
> Maxime