ASoC: UDA134X Codec: Fix mute/unmute code mistake and add ADC/DAC power control support

There is a mistake in current uda134x_mute function: mute_reg has been changed in line 162 or line 164, so uda134x_write should write "mute_reg" but not "mute_reg & ~(1<<2)" to
UDA134X_DATA010.

Besides, because there is no DAPM configuration for uda134x, when system starts up, snd_soc_int_card calls snd_soc_dapm_new_widgets, and snd_soc_dapm_new_widgets calls dapm_power_widgets. In function dapm_power_widgets, codec->dapm_widgets has no list entry, so sys_power retains it's original value zero. Then snd_soc_dapm_set_bias_level(socdev, SND_SOC_BIAS_PREPARE) and snd_soc_dapm_set_bias_level(socdev, SND_SOC_BIAS_STANDBY) are called sequentially. Finally, the uda134x codec goes to STANDBY mode, so the ADC/DAC power control bits of UDA134X_STATUS1 register keeps the 0 value.

UDA134X has no trigger function currently, and the ADC/DAC power control bits of UDA134X_STATUS1 register are not exported in uda1341_snd_controls, so there's no way to enable the ADC/DAC power control bits when you want to use the codec. When playing with aplay or recording with arecord, there is no sound output or no wave input.

I have added the uda134x_trigger function, which turns on/off the ADC/DAC power control bits according the substream type and the command. I also exported the ADC/DAC power control bits to uda1341_snd_controls, so we can also turn these bits on/off as we need via a mixer tool like alsamixer.

The patch created against linux-2.6.31-rc3.
Tested on a s3c2440 development board with UDA1341TS codec.

@@ -163,7 +163,7 @@
 	else
 		mute_reg &= ~(1<<2);
 
-	uda134x_write(codec, UDA134X_DATA010, mute_reg & ~(1<<2));
+	uda134x_write(codec, UDA134X_DATA010, mute_reg);
 
 	return 0;
 }

@@ -339,6 +339,38 @@
 	return 0;
 }
 
+static int uda134x_trigger(struct snd_pcm_substream *substream,
+			   int cmd, struct snd_soc_dai *dai)
+{
+	int stream = substream->stream;
+	struct snd_soc_codec *codec = dai->codec;
+	u8 power_ctrl_reg = uda134x_read_reg_cache(codec, UDA134X_STATUS1);
+
+	pr_debug("%s stream: %d, cmd: %d\n", __func__, stream, cmd);
+
+	switch (cmd) {
+	case SNDRV_PCM_TRIGGER_START:
+		if (stream == SNDRV_PCM_STREAM_PLAYBACK)
+			power_ctrl_reg |= (1<<0);
+		else
+			power_ctrl_reg |= (1<<1);
+		break;
+
+	case SNDRV_PCM_TRIGGER_STOP:
+		if (stream == SNDRV_PCM_STREAM_PLAYBACK)
+			power_ctrl_reg &= ~(1<<0);
+		else
+			power_ctrl_reg &= ~(1<<1);
+		break;
+
+	default:
+		return 0;
+	}
+
+	uda134x_write(codec, UDA134X_STATUS1, power_ctrl_reg);
+	return 0;
+}
+
 static int uda134x_set_bias_level(struct snd_soc_codec *codec,
 				  enum snd_soc_bias_level level)
 {

@@ -416,6 +448,8 @@
 SOC_SINGLE("ADC Polarity Switch", UDA134X_STATUS1, 4, 1, 0),
 SOC_SINGLE("DAC Polarity Switch", UDA134X_STATUS1, 3, 1, 0),
 SOC_SINGLE("Double Speed Playback Switch", UDA134X_STATUS1, 2, 1, 0),
+SOC_SINGLE("ADC Power Switch", UDA134X_STATUS1, 1, 1, 0),
+SOC_SINGLE("DAC Power Switch", UDA134X_STATUS1, 0, 1, 0),
 SOC_SINGLE("DC Filter Enable Switch", UDA134X_STATUS0, 0, 1, 0),
 };
 

@@ -438,6 +472,7 @@
 	.digital_mute	= uda134x_mute,
 	.set_sysclk	= uda134x_set_dai_sysclk,
 	.set_fmt	= uda134x_set_dai_fmt,
+	.trigger	= uda134x_trigger,
 };
 
 struct snd_soc_dai uda134x_dai = {

This is OK but please split it out into a separate patch and resend since...

...this isn't, partly because I can't really follow what you're trying to do here. It looks awfully like you're trying to implement DAPM support - certainly things like adding the switches for DAC and ADC power shouldn't be there, they should be automatically managed in the drivers.

Resend the patch for the mistaken mute/unmute code.

@@ -163,7 +163,7 @@
 	else
 		mute_reg &= ~(1<<2);
 
-	uda134x_write(codec, UDA134X_DATA010, mute_reg & ~(1<<2));
+	uda134x_write(codec, UDA134X_DATA010, mute_reg);
 
 	return 0;
 }

I've applied this after copying the changelog from your previous patch.

When I play a wav file using aplay, I found there's no sound. After some debug work, I noticed that the DAC power control bit keeps zero since the system started up.

DAC power control bit should be set when the codec is about to excute a playback task. But current code doesn't set this bit on before playback. So I do the work in the codec's trigger operation.

So is the wave recording. There's no code to set the ADC power control bit on when to record.

Yes, it's true that we should not expose the power control to the end user. So I should the remove the two lines in the last patch.

Please try the patch below.

commit 6294523841a61cd1ad24f220daf6e73e9cc83184

Author: Mark Brown <[email protected]>
Date:   Mon Aug 17 11:55:38 2009 +0100

ASoC: Fix handling of bias levels for non-DAPM codecs

If the system doesn't have any DAPM widgets then we can't use their state to check if the bias level for the codec should be up.

index c68c204..4a21a5e 100644
@@ -966,6 +966,22 @@ static int dapm_power_widgets(struct snd_soc_codec *codec, int event)
 		}
 	}
 
+	/* If there are no DAPM widgets then try to figure out power from the
+	 * event type.
+	 */
+	if (list_empty(&codec->dapm_widgets)) {
+		switch (event) {
+		case SND_SOC_DAPM_STREAM_START:
+		case SND_SOC_DAPM_STREAM_RESUME:
+			sys_power = 1;
+			break;
+		case SND_SOC_DAPM_STREAM_NOP:
+			sys_power == codec->bias_level != SND_SOC_BIAS_STANDBY;
+		default:
+			break;
+		}
+	}
+
 	/* If we're changing to all on or all off then prepare */
 	if ((sys_power && codec->bias_level == SND_SOC_BIAS_STANDBY) ||
 	    (!sys_power && codec->bias_level == SND_SOC_BIAS_ON)) {

Yes.

Thank you for your remainder.
Using -p1 regenerate the patch.

@@ -163,7 +163,7 @@
 	else
 		mute_reg &= ~(1<<2);
 
-	uda134x_write(codec, UDA134X_DATA010, mute_reg & ~(1<<2));
+	uda134x_write(codec, UDA134X_DATA010, mute_reg);
 
 	return 0;
 }