On Wed, Jun 29, 2022 at 10:06:43AM +0100, Srinivas Kandagatla wrote:Yes my bad..
+static int wsa_dev_mode_put(struct snd_kcontrol *kcontrol,
+ struct snd_ctl_elem_value *ucontrol)
+{
+ struct snd_soc_component *component = snd_soc_kcontrol_component(kcontrol);
+ struct wsa883x_priv *wsa883x = snd_soc_component_get_drvdata(component);
+
+ if (wsa883x->dev_mode == ucontrol->value.enumerated.item[0])
+ return 0;
+
+ wsa883x->dev_mode = ucontrol->value.enumerated.item[0];
+
+ return 0;
+}
This isn't returning 1 when the value changes so will miss generating
events, please test the driver with mixer-test.
Sounds good,
+ switch (event) {
+ case SND_SOC_DAPM_POST_PMU:
+ if (wsa883x->dev_mode == RECEIVER) {
+ snd_soc_component_write_field(component, WSA883X_CDC_PATH_MODE,
+ WSA883X_RXD_MODE_MASK,
+ WSA883X_RXD_MODE_HIFI);
+ snd_soc_component_write_field(component, WSA883X_SPKR_PWM_CLK_CTL,
+ WSA883X_SPKR_PWM_FREQ_SEL_MASK,
+ WSA883X_SPKR_PWM_FREQ_F600KHZ);
+ snd_soc_component_write_field(component, WSA883X_DRE_CTL_0,
+ WSA883X_DRE_PROG_DELAY_MASK, 0x0);
+ } else if (wsa883x->dev_mode == SPEAKER) {
This looks like it'd be better written as a switch statement.
Okay, will fix all these in next spin.
+static const struct snd_kcontrol_new wsa883x_snd_controls[] = {
+ SOC_SINGLE_RANGE_TLV("PA Gain", WSA883X_DRE_CTL_1, 1,
+ 0x0, 0x1f, 1, pa_gain),
Volume controls should end in Volume, mixer-test should also have caught
this.