From: Marcus Cooper <[email protected]>
We have a number of flags used to identify the functionality
of the IP block found on the sun8i-h3 and later devices. As it
is only neccessary to identify this new block then replace
these flags with just one.
Signed-off-by: Marcus Cooper <[email protected]>
---
sound/soc/sunxi/sun4i-i2s.c | 23 +++++++----------------
1 file changed, 7 insertions(+), 16 deletions(-)
diff --git a/sound/soc/sunxi/sun4i-i2s.c b/sound/soc/sunxi/sun4i-i2s.c
index e2961d8f6e8c..329883750d6f 100644
--- a/sound/soc/sunxi/sun4i-i2s.c
+++ b/sound/soc/sunxi/sun4i-i2s.c
@@ -119,10 +119,7 @@
*
* @has_reset: SoC needs reset deasserted.
* @has_slave_select_bit: SoC has a bit to enable slave mode.
- * @has_fmt_set_lrck_period: SoC requires lrclk period to be set.
- * @has_chcfg: tx and rx slot number need to be set.
- * @has_chsel_tx_chen: SoC requires that the tx channels are enabled.
- * @has_chsel_offset: SoC uses offset for selecting dai operational mode.
+ * @is_h3_i2s_based: This block is similiar to what is found on the h3.
* @reg_offset_txdata: offset of the tx fifo.
* @sun4i_i2s_regmap: regmap config to use.
* @mclk_offset: Value by which mclkdiv needs to be adjusted.
@@ -143,10 +140,7 @@
struct sun4i_i2s_quirks {
bool has_reset;
bool has_slave_select_bit;
- bool has_fmt_set_lrck_period;
- bool has_chcfg;
- bool has_chsel_tx_chen;
- bool has_chsel_offset;
+ bool is_h3_i2s_based;
unsigned int reg_offset_txdata; /* TX FIFO */
const struct regmap_config *sun4i_i2s_regmap;
unsigned int mclk_offset;
@@ -340,7 +334,7 @@ static int sun4i_i2s_set_clk_rate(struct snd_soc_dai *dai,
regmap_field_write(i2s->field_clkdiv_mclk_en, 1);
/* Set sync period */
- if (i2s->variant->has_fmt_set_lrck_period)
+ if (i2s->variant->is_h3_i2s_based)
regmap_update_bits(i2s->regmap, SUN4I_I2S_FMT0_REG,
SUN8I_I2S_FMT0_LRCK_PERIOD_MASK,
SUN8I_I2S_FMT0_LRCK_PERIOD(32));
@@ -366,7 +360,7 @@ static int sun4i_i2s_hw_params(struct snd_pcm_substream *substream,
return -EINVAL;
}
- if (i2s->variant->has_chcfg) {
+ if (i2s->variant->is_h3_i2s_based) {
regmap_update_bits(i2s->regmap, SUN8I_I2S_CHAN_CFG_REG,
SUN8I_I2S_CHAN_CFG_TX_SLOT_NUM_MASK,
SUN8I_I2S_CHAN_CFG_TX_SLOT_NUM(channels));
@@ -386,7 +380,7 @@ static int sun4i_i2s_hw_params(struct snd_pcm_substream *substream,
regmap_field_write(i2s->field_rxchansel,
SUN4I_I2S_CHAN_SEL(params_channels(params)));
- if (i2s->variant->has_chsel_tx_chen)
+ if (i2s->variant->is_h3_i2s_based)
regmap_update_bits(i2s->regmap, SUN8I_I2S_TX_CHAN_SEL_REG,
SUN8I_I2S_TX_CHAN_EN_MASK,
SUN8I_I2S_TX_CHAN_EN(channels));
@@ -449,7 +443,7 @@ static int sun4i_i2s_set_fmt(struct snd_soc_dai *dai, unsigned int fmt)
return -EINVAL;
}
- if (i2s->variant->has_chsel_offset) {
+ if (i2s->variant->is_h3_i2s_based) {
/*
* offset being set indicates that we're connected to an i2s
* device, however offset is only used on the sun8i block and
@@ -942,10 +936,7 @@ static const struct sun4i_i2s_quirks sun8i_h3_i2s_quirks = {
.mclk_offset = 1,
.bclk_offset = 2,
.fmt_offset = 3,
- .has_fmt_set_lrck_period = true,
- .has_chcfg = true,
- .has_chsel_tx_chen = true,
- .has_chsel_offset = true,
+ .is_h3_i2s_based = true,
.field_clkdiv_mclk_en = REG_FIELD(SUN4I_I2S_CLK_DIV_REG, 8, 8),
.field_fmt_wss = REG_FIELD(SUN4I_I2S_FMT0_REG, 0, 2),
.field_fmt_sr = REG_FIELD(SUN4I_I2S_FMT0_REG, 4, 6),
--
2.21.0
On Mon, Jun 03, 2019 at 07:47:30PM +0200, [email protected] wrote:
> From: Marcus Cooper <[email protected]>
>
> We have a number of flags used to identify the functionality
> of the IP block found on the sun8i-h3 and later devices. As it
> is only neccessary to identify this new block then replace
> these flags with just one.
>
> Signed-off-by: Marcus Cooper <[email protected]>
This carries exactly the same meaning than the compatible, so this is
entirely redundant.
The more I think of it, the more I fell like we should have function
pointers instead, and have hooks to deal with these kind of things.
I've been working a lot on that driver recently, and there's some many
parameters and regmap_fields that it becomes pretty hard to work on.
Maxime
--
Maxime Ripard, Bootlin
Embedded Linux and Kernel engineering
https://bootlin.com
On Tue, 4 Jun 2019 at 09:46, Maxime Ripard <[email protected]> wrote:
>
> On Mon, Jun 03, 2019 at 07:47:30PM +0200, [email protected] wrote:
> > From: Marcus Cooper <[email protected]>
> >
> > We have a number of flags used to identify the functionality
> > of the IP block found on the sun8i-h3 and later devices. As it
> > is only neccessary to identify this new block then replace
> > these flags with just one.
> >
> > Signed-off-by: Marcus Cooper <[email protected]>
>
> This carries exactly the same meaning than the compatible, so this is
> entirely redundant.
>
> The more I think of it, the more I fell like we should have function
> pointers instead, and have hooks to deal with these kind of things.
>
> I've been working a lot on that driver recently, and there's some many
> parameters and regmap_fields that it becomes pretty hard to work on.
Hi Maxime,
let's sync with what you're doing as you're more lightly to see it
through to delivery!
I was trying to clean up the driver as some of this seemed a bit unnecessary,
hooks sounds like the way forward,
CK
>
> Maxime
>
> --
> Maxime Ripard, Bootlin
> Embedded Linux and Kernel engineering
> https://bootlin.com