On Thu, May 02, 2019 at 11:39:24AM +0200, Clément Péron wrote:
> > > @@ -169,16 +181,25 @@ struct sun4i_spdif_dev {
> > > struct snd_soc_dai_driver cpu_dai_drv;
> > > struct regmap *regmap;
> > > struct snd_dmaengine_dai_dma_data dma_params_tx;
> > > + const struct sun4i_spdif_quirks *quirks;
> >
> > I guess this will generate a warning since the structure hasn't been
> > defined yet?
>
> It's a pointer to a structure so no warning from the compiler.
Damn, I was convinced just declaring a pointer to a structure would
result to a gcc warning. Nevermind then.
> > > @@ -405,22 +426,26 @@ static struct snd_soc_dai_driver sun4i_spdif_dai = {
> > > .name = "spdif",
> > > };
> > >
> > > -struct sun4i_spdif_quirks {
> > > - unsigned int reg_dac_txdata; /* TX FIFO offset for DMA config */
> > > - bool has_reset;
> > > -};
> > > -
> > > static const struct sun4i_spdif_quirks sun4i_a10_spdif_quirks = {
> > > .reg_dac_txdata = SUN4I_SPDIF_TXFIFO,
> > > + .reg_fctl_ftx = SUN4I_SPDIF_FCTL_FTX,
> > > };
> > >
> > > static const struct sun4i_spdif_quirks sun6i_a31_spdif_quirks = {
> > > .reg_dac_txdata = SUN4I_SPDIF_TXFIFO,
> > > + .reg_fctl_ftx = SUN4I_SPDIF_FCTL_FTX,
> > > .has_reset = true,
> > > };
> > >
> > > static const struct sun4i_spdif_quirks sun8i_h3_spdif_quirks = {
> > > .reg_dac_txdata = SUN8I_SPDIF_TXFIFO,
> > > + .reg_fctl_ftx = SUN4I_SPDIF_FCTL_FTX,
> > > + .has_reset = true,
> > > +};
> > >
> > > +static const struct sun4i_spdif_quirks sun50i_h6_spdif_quirks = {
> > > + .reg_dac_txdata = SUN8I_SPDIF_TXFIFO,
> > > + .reg_fctl_ftx = SUN50I_H6_SPDIF_FCTL_FTX,
> > > .has_reset = true,
> >
> > The reg_dac_txdata and reg_fctl_ftx changes here should also be part
> > of a separate patch.
>
> You mean the reg_fctl_ftx quirk and the H6 introduction should be split ?
Yep
Maxime
--
Maxime Ripard, Bootlin
Embedded Linux and Kernel engineering
https://bootlin.com
Attachment:
signature.asc
Description: PGP signature
