El Thu, Feb 15, 2018 at 09:42:01AM +0200 Peter Ujfalusi ha dit:
On 2018-02-15 01:51, Matthias Kaehlcke wrote:
On some systems a delay is needed after switching on the clocks, to allow the DMIC output to stabilize and avoid a popping noise at the beginning of the recording. Add the optional device tree property 'wakeup-delay-ms' and apply the specified delay in the new dmic_daiops_prepare().
Signed-off-by: Matthias Kaehlcke mka@chromium.org
.../devicetree/bindings/sound/dmic.txt | 2 + sound/soc/codecs/dmic.c | 54 ++++++++++++++----- 2 files changed, 42 insertions(+), 14 deletions(-)
diff --git a/Documentation/devicetree/bindings/sound/dmic.txt b/Documentation/devicetree/bindings/sound/dmic.txt index 54c8ef6498a8..de741c6609d0 100644 --- a/Documentation/devicetree/bindings/sound/dmic.txt +++ b/Documentation/devicetree/bindings/sound/dmic.txt @@ -7,10 +7,12 @@ Required properties:
Optional properties:
- dmicen-gpios: GPIO specifier for dmic to control start and stop
- wakeup-delay-ms: Delay (in ms) after enabling the DMIC
Example node:
dmic_codec: dmic@0 { compatible = "dmic-codec"; dmicen-gpios = <&gpio4 3 GPIO_ACTIVE_HIGH>;
wakeup-delay-ms <50>;diff --git a/sound/soc/codecs/dmic.c b/sound/soc/codecs/dmic.c index b88a1ee66f80..11f6abf11074 100644 --- a/sound/soc/codecs/dmic.c +++ b/sound/soc/codecs/dmic.c @@ -19,6 +19,7 @@
*/
+#include <linux/delay.h> #include <linux/gpio.h> #include <linux/gpio/consumer.h> #include <linux/platform_device.h> @@ -29,24 +30,38 @@ #include <sound/soc.h> #include <sound/soc-dapm.h>
+struct dmic {
- struct gpio_desc *gpio_en;
- int wakeup_delay;
+};
+static int dmic_daiops_prepare(struct snd_pcm_substream *substream,
struct snd_soc_dai *dai)+{
- struct dmic *dmic = snd_soc_dai_get_drvdata(dai);
- if (dmic->gpio_en)
gpiod_set_value(dmic->gpio_en, 1);You don't need the 'if (dmic->gpio_en)'
True, personally I think the if statement makes it clearer that the GPIO is optional. I'm fine with removing it if the general sense is that it's just noise.
- if (dmic->wakeup_delay)
msleep(dmic->wakeup_delay);- return 0;
+}
static int dmic_daiops_trigger(struct snd_pcm_substream *substream, int cmd, struct snd_soc_dai *dai) {
- struct gpio_desc *dmic_en = snd_soc_dai_get_drvdata(dai);
- struct dmic *dmic = snd_soc_dai_get_drvdata(dai);
- if (!dmic_en)
if (!dmic->gpio_en) return 0;
switch (cmd) {
- case SNDRV_PCM_TRIGGER_START:
- case SNDRV_PCM_TRIGGER_RESUME:
- case SNDRV_PCM_TRIGGER_PAUSE_RELEASE:
gpiod_set_value(dmic_en, 1);break;gpiod_set_value(dmic_en, 0);
gpiod_set_value(dmic->gpio_en, 0);What if instead of this trickery we try to handle the gpio/delay via DAPM? SND_SOC_DAPM_AIF_OUT_E() might be just what we need?
We could remove the trigger, and will have no need for the prepare callback either..
SND_SOC_DAPM_AIF_OUT_E() looks promising, thanks for the pointer!
@@ -54,6 +69,7 @@ static int dmic_daiops_trigger(struct snd_pcm_substream *substream, }
static const struct snd_soc_dai_ops dmic_dai_ops = {
- .prepare = dmic_daiops_prepare, .trigger = dmic_daiops_trigger,
};
@@ -73,14 +89,24 @@ static struct snd_soc_dai_driver dmic_dai = {
static int dmic_codec_probe(struct snd_soc_codec *codec) {
- struct gpio_desc *dmic_en;
- struct dmic *dmic;
- int err;
- dmic = devm_kzalloc(codec->dev, sizeof(*dmic), GFP_KERNEL);
- if (!dmic)
return -ENOMEM;- dmic->gpio_en = devm_gpiod_get_optional(codec->dev,
"dmicen", GPIOD_OUT_LOW);- if (IS_ERR(dmic->gpio_en))
return PTR_ERR(dmic->gpio_en);
- dmic_en = devm_gpiod_get_optional(codec->dev,
"dmicen", GPIOD_OUT_LOW);- if (IS_ERR(dmic_en))
return PTR_ERR(dmic_en);
- err = device_property_read_u32(codec->dev, "wakeup-delay-ms",
&dmic->wakeup_delay);- if (err && (err != -EINVAL))
return err;
- snd_soc_codec_set_drvdata(codec, dmic_en);
snd_soc_codec_set_drvdata(codec, dmic);
return 0;
}
- Péter
Texas Instruments Finland Oy, Porkkalankatu 22, 00180 Helsinki. Y-tunnus/Business ID: 0615521-4. Kotipaikka/Domicile: Helsinki