Skip to contentSkip to conversation

ASoC: sgtl5000: Revert "ASoC: sgtl5000: Fix of unmute outputs on probe"

PATCH In review Alison Wang 6 participants 16 messages Dec 12, 2019 – Mar 20, 2020

Subsystems (3)
lore
Share & tools
lore
#0 lore raw

PATCH 1/1 ASoC: sgtl5000: Revert "ASoC: sgtl5000: Fix of unmute outputs on probe"

This reverts commit 631bc8f0134ae9620d86a96b8c5f9445d91a2dca.

In commit 631bc8f0134a, snd_soc_component_update_bits is used instead of snd_soc_component_write. Although EN_HP_ZCD and EN_ADC_ZCD are enabled in ANA_CTRL register, MUTE_LO, MUTE_HP and MUTE_ADC bits are remained as the default value. It causes LO, HP and ADC are all muted after driver probe.

The patch is to revert this commit, snd_soc_component_write is still used and MUTE_LO, MUTE_HP and MUTE_ADC are set as zero (unmuted).

---
 sound/soc/codecs/sgtl5000.c | 6 +++---
 1 file changed, 3 insertions(+), 3 deletions(-)
sound/soc/codecs/sgtl5000.c +3 -3
index aa1f963..0b35fca 100644
@@ -1458,7 +1458,6 @@ static int sgtl5000_probe(struct snd_soc_component *component)
 	int ret;
 	u16 reg;
 	struct sgtl5000_priv *sgtl5000 = snd_soc_component_get_drvdata(component);
-	unsigned int zcd_mask = SGTL5000_HP_ZCD_EN | SGTL5000_ADC_ZCD_EN;
 
 	/* power up sgtl5000 */
 	ret = sgtl5000_set_power_regs(component);
@@ -1486,8 +1485,9 @@ static int sgtl5000_probe(struct snd_soc_component *component)
 	       0x1f);
 	snd_soc_component_write(component, SGTL5000_CHIP_PAD_STRENGTH, reg);
 
-	snd_soc_component_update_bits(component, SGTL5000_CHIP_ANA_CTRL,
-		zcd_mask, zcd_mask);
+	snd_soc_component_write(component, SGTL5000_CHIP_ANA_CTRL,
+			SGTL5000_HP_ZCD_EN |
+			SGTL5000_ADC_ZCD_EN);
 
 	snd_soc_component_update_bits(component, SGTL5000_CHIP_MIC_CTRL,
 			SGTL5000_BIAS_R_MASK,
Signature · 1 line
2.9.5
#1 lore raw

Alison, could you please explain, what reason to revert this fix? All these blocks have their own control and unmute on demand.
Why do you stand on unconditional unmuting of outputs and ADC on driver probing?

Quoted 44 lines · Alison Wang

This reverts commit 631bc8f0134ae9620d86a96b8c5f9445d91a2dca.

In commit 631bc8f0134a, snd_soc_component_update_bits is used instead of snd_soc_component_write. Although EN_HP_ZCD and EN_ADC_ZCD are enabled in ANA_CTRL register, MUTE_LO, MUTE_HP and MUTE_ADC bits are remained as the default value. It causes LO, HP and ADC are all muted after driver probe.

The patch is to revert this commit, snd_soc_component_write is still used and MUTE_LO, MUTE_HP and MUTE_ADC are set as zero (unmuted).

---
 sound/soc/codecs/sgtl5000.c | 6 +++---
 1 file changed, 3 insertions(+), 3 deletions(-)
sound/soc/codecs/sgtl5000.c +3 -4
index aa1f963..0b35fca 100644
@@ -1458,7 +1458,6 @@ static int sgtl5000_probe(struct snd_soc_component *component)
        int ret;
        u16 reg;
        struct sgtl5000_priv *sgtl5000 = snd_soc_component_get_drvdata(component);
-       unsigned int zcd_mask = SGTL5000_HP_ZCD_EN | SGTL5000_ADC_ZCD_EN;
 
        /* power up sgtl5000 */
        ret = sgtl5000_set_power_regs(component);
@@ -1486,8 +1485,9 @@ static int sgtl5000_probe(struct snd_soc_component *component)
               0x1f);
        snd_soc_component_write(component, SGTL5000_CHIP_PAD_STRENGTH, reg);
 
-       snd_soc_component_update_bits(component, SGTL5000_CHIP_ANA_CTRL,
-               zcd_mask, zcd_mask);
+       snd_soc_component_write(component, SGTL5000_CHIP_ANA_CTRL,
+                       SGTL5000_HP_ZCD_EN |
+                       SGTL5000_ADC_ZCD_EN);
 
        snd_soc_component_update_bits(component, SGTL5000_CHIP_MIC_CTRL,
                        SGTL5000_BIAS_R_MASK,
--

2.9.5

Signature · 10 lines
Best regards
Oleksandr Suvorov

Toradex AG
Altsagenstrasse 5 | 6048 Horw/Luzern | Switzerland | T: +41 41 500
4800 (main line)
_______________________________________________
Alsa-devel mailing list
[email protected]
https://mailman.alsa-project.org/mailman/listinfo/alsa-devel
#2 lore raw

Hi, Oleksandr,

I think I have explained the reason in my email which sent to you yesterday. I attached it too. According to my test on the boards, MUTE_LO, MUTE_HP and MUTE_ADC are all muted when playing the sound.

Could you give me some ideas about this issue?

Best Regards,I
Alison Wang

Quoted 75 lines

-----Original Message-----
From: Oleksandr Suvorov <[email protected]>
Sent: Thursday, December 12, 2019 5:11 PM
To: Alison Wang <[email protected]>
Cc: Marcel Ziswiler <[email protected]>; Igor Opaniuk <[email protected]>; [email protected]; [email protected]; [email protected]; Oleksandr Suvorov <[email protected]>; [email protected]; [email protected]
Subject: [EXT] Re: [PATCH] ASoC: sgtl5000: Revert "ASoC: sgtl5000: Fix of unmute outputs on probe"

Caution: EXT Email

Alison, could you please explain, what reason to revert this fix? All these blocks have their own control and unmute on demand.
Why do you stand on unconditional unmuting of outputs and ADC on driver probing?

sound/soc/codecs/sgtl5000.c static int sgtl5000_probe(struct

Quoted 22 lines · Alison Wang

This reverts commit 631bc8f0134ae9620d86a96b8c5f9445d91a2dca.

In commit 631bc8f0134a, snd_soc_component_update_bits is used instead of snd_soc_component_write. Although EN_HP_ZCD and EN_ADC_ZCD are enabled in ANA_CTRL register, MUTE_LO, MUTE_HP and MUTE_ADC bits are remained as the default value. It causes LO, HP and ADC are all muted after driver probe.

The patch is to revert this commit, snd_soc_component_write is still used and MUTE_LO, MUTE_HP and MUTE_ADC are set as zero (unmuted).

---
 sound/soc/codecs/sgtl5000.c | 6 +++---
 1 file changed, 3 insertions(+), 3 deletions(-)
sound/soc/codecs/sgtl5000.c +0 -0
index aa1f963..0b35fca 100644
@@ -1458,7 +1458,6 @@ static int sgtl5000_probe(struct

snd_soc_component *component)

sound/soc/codecs/sgtl5000.c static int sgtl5000_probe(struct snd_soc_component *component)

Quoted 3 lines
sound/soc/codecs/sgtl5000.c +0 -0
        int ret;
        u16 reg;
        struct sgtl5000_priv *sgtl5000 =

snd_soc_component_get_drvdata(component);

sound/soc/codecs/sgtl5000.c static int sgtl5000_probe(struct · line 1461

Quoted 1 line
sound/soc/codecs/sgtl5000.c +0 -1
-       unsigned int zcd_mask = SGTL5000_HP_ZCD_EN |

SGTL5000_ADC_ZCD_EN;

sound/soc/codecs/sgtl5000.c static int sgtl5000_probe(struct

Quoted 4 lines
sound/soc/codecs/sgtl5000.c +0 -0
        /* power up sgtl5000 */
        ret = sgtl5000_set_power_regs(component);
@@ -1486,8 +1485,9 @@ static int sgtl5000_probe(struct

snd_soc_component *component)

sound/soc/codecs/sgtl5000.c static int sgtl5000_probe(struct · line 1486

Quoted 2 lines
sound/soc/codecs/sgtl5000.c +0 -0
               0x1f);
        snd_soc_component_write(component,

SGTL5000_CHIP_PAD_STRENGTH,

sound/soc/codecs/sgtl5000.c static int sgtl5000_probe(struct · line 1490

Quoted 3 lines
sound/soc/codecs/sgtl5000.c +0 -1
 reg);
 
-       snd_soc_component_update_bits(component,

SGTL5000_CHIP_ANA_CTRL,

sound/soc/codecs/sgtl5000.c static int sgtl5000_probe(struct · line 1489

Quoted 2 lines
sound/soc/codecs/sgtl5000.c +1 -1
-               zcd_mask, zcd_mask);
+       snd_soc_component_write(component,

SGTL5000_CHIP_ANA_CTRL,

sound/soc/codecs/sgtl5000.c static int sgtl5000_probe(struct snd_soc_component *component)

Quoted 4 lines
sound/soc/codecs/sgtl5000.c +2 -0
+                       SGTL5000_HP_ZCD_EN |
+                       SGTL5000_ADC_ZCD_EN);
 
        snd_soc_component_update_bits(component,

SGTL5000_CHIP_MIC_CTRL,

Quoted 4 lines
                        SGTL5000_BIAS_R_MASK,

--
2.9.5

Quoted 7 lines

--
Best regards
Oleksandr Suvorov

Toradex AG
Altsagenstrasse 5 | 6048 Horw/Luzern | Switzerland | T: +41 41 500 4800 (main line)

Hi, Oleksandr,

I have a question about this patch. Could you give me some ideas? Thanks.

In this patch, snd_soc_component_update_bits is used instead of snd_soc_component_write. Although EN_HP_ZCD and EN_ADC_ZCD are enabled in ANA_CTRL register, MUTE_LO, MUTE_HP and MUTE_ADC bits are remained as the default value. It causes LO, HP and ADC are all muted after driver probe. BTW, in the original code, snd_soc_component_write is used and MUTE_LO, MUTE_HP and MUTE_ADC are all set as zero (unmute).

I saw the above phenomenon on the latest linux-next. For LO and HP are muted, no sound can be heard.

Best Regards,
Alison Wang

Quoted 55 lines

Subject: [v6,5/6] ASoC: sgtl5000: Fix of unmute outputs on probe

To enable "zero cross detect" for ADC/HP, change HP_ZCD_EN/ADC_ZCD_EN bits only instead of writing the whole CHIP_ANA_CTRL register.

---

Changes in v6: None
Changes in v5: None
Changes in v4: None
Changes in v3: None
Changes in v2:
- Fix patch formatting

 sound/soc/codecs/sgtl5000.c | 6 +++---
 1 file changed, 3 insertions(+), 3 deletions(-)
sound/soc/codecs/sgtl5000.c index renamed from sound/soc/codecs/sgtl5000.c +0 -0

b65232521ea8..23f4ae2f0723 100644

sound/soc/codecs/sgtl5000.c +1 -0
@@ -1453,6 +1453,7 @@  static int sgtl5000_probe(struct
 snd_soc_component *component)
 	int ret;
 	u16 reg;
 	struct sgtl5000_priv *sgtl5000 =
 snd_soc_component_get_drvdata(component);
+	unsigned int zcd_mask = SGTL5000_HP_ZCD_EN |
 SGTL5000_ADC_ZCD_EN;
sound/soc/codecs/sgtl5000.c +2 -3
 	/* power up sgtl5000 */
 	ret = sgtl5000_set_power_regs(component);
@@ -1480,9 +1481,8 @@  static int sgtl5000_probe(struct
 snd_soc_component *component)
 	       0x1f);
 	snd_soc_component_write(component, SGTL5000_CHIP_PAD_STRENGTH,
 reg);
 
-	snd_soc_component_write(component, SGTL5000_CHIP_ANA_CTRL,
-			SGTL5000_HP_ZCD_EN |
-			SGTL5000_ADC_ZCD_EN);
+	snd_soc_component_update_bits(component,
 SGTL5000_CHIP_ANA_CTRL,
+		zcd_mask, zcd_mask);

snd_soc_component_update_bits(component,
SGTL5000_CHIP_MIC_CTRL,
SGTL5000_BIAS_R_MASK,

#3 lore raw

Hi, Oleksandr,

In your patch, outputs and ADC are muted on driver probing. I don't know when and where they can be unmuted in the kernel before playing in the application.

Could you please give me some ideas?

Best Regards,
Alison Wang

Quoted 110 lines

-----Original Message-----
From: Alison Wang
Sent: Thursday, December 12, 2019 5:25 PM
To: Oleksandr Suvorov <[email protected]>
Cc: Marcel Ziswiler <[email protected]>; Igor Opaniuk <[email protected]>; [email protected]; [email protected]; [email protected]; [email protected]; [email protected]
Subject: RE: [EXT] Re: [PATCH] ASoC: sgtl5000: Revert "ASoC: sgtl5000: Fix of unmute outputs on probe"

Hi, Oleksandr,

I think I have explained the reason in my email which sent to you yesterday. I attached it too.
According to my test on the boards, MUTE_LO, MUTE_HP and MUTE_ADC are all muted when playing the sound.

Could you give me some ideas about this issue?

Best Regards,I
Alison Wang

Quoted 7 lines

-----Original Message-----
From: Oleksandr Suvorov <[email protected]>
Sent: Thursday, December 12, 2019 5:11 PM
To: Alison Wang <[email protected]>
Cc: Marcel Ziswiler <[email protected]>; Igor Opaniuk <[email protected]>; [email protected]; [email protected]; [email protected]; Oleksandr Suvorov

<[email protected]>;

Quoted 7 lines

[email protected]; [email protected]
Subject: [EXT] Re: [PATCH] ASoC: sgtl5000: Revert "ASoC: sgtl5000: Fix of unmute outputs on probe"

Caution: EXT Email

Alison, could you please explain, what reason to revert this fix? All these

blocks

Quoted 4 lines

have their own control and unmute on demand.
Why do you stand on unconditional unmuting of outputs and ADC on driver probing?

Quoted 8 lines
Quoted 6 lines · Alison Wang

This reverts commit 631bc8f0134ae9620d86a96b8c5f9445d91a2dca.

In commit 631bc8f0134a, snd_soc_component_update_bits is used instead of snd_soc_component_write. Although EN_HP_ZCD and EN_ADC_ZCD are enabled in ANA_CTRL register, MUTE_LO, MUTE_HP and MUTE_ADC bits

are

Quoted 56 lines

sound/soc/codecs/sgtl5000.c static int sgtl5000_probe(struct

Quoted 16 lines

remained as the default value. It causes LO, HP and ADC are all muted after driver probe.

The patch is to revert this commit, snd_soc_component_write is still used and MUTE_LO, MUTE_HP and MUTE_ADC are set as zero (unmuted).

---
 sound/soc/codecs/sgtl5000.c | 6 +++---
 1 file changed, 3 insertions(+), 3 deletions(-)
sound/soc/codecs/sgtl5000.c +0 -0
index aa1f963..0b35fca 100644
@@ -1458,7 +1458,6 @@ static int sgtl5000_probe(struct

snd_soc_component *component)

sound/soc/codecs/sgtl5000.c static int sgtl5000_probe(struct snd_soc_component *component)

Quoted 3 lines
sound/soc/codecs/sgtl5000.c +0 -0
        int ret;
        u16 reg;
        struct sgtl5000_priv *sgtl5000 =

snd_soc_component_get_drvdata(component);

sound/soc/codecs/sgtl5000.c static int sgtl5000_probe(struct · line 1461

Quoted 1 line
sound/soc/codecs/sgtl5000.c +0 -1
-       unsigned int zcd_mask = SGTL5000_HP_ZCD_EN |

SGTL5000_ADC_ZCD_EN;

sound/soc/codecs/sgtl5000.c static int sgtl5000_probe(struct

Quoted 4 lines
sound/soc/codecs/sgtl5000.c +0 -0
        /* power up sgtl5000 */
        ret = sgtl5000_set_power_regs(component);
@@ -1486,8 +1485,9 @@ static int sgtl5000_probe(struct

snd_soc_component *component)

sound/soc/codecs/sgtl5000.c static int sgtl5000_probe(struct · line 1486

Quoted 2 lines
sound/soc/codecs/sgtl5000.c +0 -0
               0x1f);
        snd_soc_component_write(component,

SGTL5000_CHIP_PAD_STRENGTH,

sound/soc/codecs/sgtl5000.c static int sgtl5000_probe(struct · line 1490

Quoted 3 lines
sound/soc/codecs/sgtl5000.c +0 -1
 reg);
 
-       snd_soc_component_update_bits(component,

SGTL5000_CHIP_ANA_CTRL,

sound/soc/codecs/sgtl5000.c static int sgtl5000_probe(struct · line 1489

Quoted 2 lines
sound/soc/codecs/sgtl5000.c +1 -1
-               zcd_mask, zcd_mask);
+       snd_soc_component_write(component,

SGTL5000_CHIP_ANA_CTRL,

sound/soc/codecs/sgtl5000.c static int sgtl5000_probe(struct snd_soc_component *component)

Quoted 4 lines
sound/soc/codecs/sgtl5000.c +2 -0
+                       SGTL5000_HP_ZCD_EN |
+                       SGTL5000_ADC_ZCD_EN);
 
        snd_soc_component_update_bits(component,

SGTL5000_CHIP_MIC_CTRL,

Quoted 4 lines
                        SGTL5000_BIAS_R_MASK,

--
2.9.5

--
Best regards
Oleksandr Suvorov

Toradex AG
Altsagenstrasse 5 | 6048 Horw/Luzern | Switzerland | T: +41 41 500 4800 (main line)

#4 lore raw

They unmute with standard sound controls:
static const struct snd_kcontrol_new sgtl5000_snd_controls[] = {
...
SOC_SINGLE("Capture Switch", SGTL5000_CHIP_ANA_CTRL, 0, 1, 1),
...
SOC_SINGLE("Headphone Playback Switch", SGTL5000_CHIP_ANA_CTRL, 4, 1, 1),
...
SOC_SINGLE("Lineout Playback Switch", SGTL5000_CHIP_ANA_CTRL, 8, 1, 1),
...

We tested this standard solution using gstreamer and standard ALSA tools like aplay, arecord and all these tools unmute needed blocks successfully.

Quoted 122 lines · Alison Wang

Hi, Oleksandr,

In your patch, outputs and ADC are muted on driver probing. I don't know when and where they can be unmuted in the kernel before playing in the application.

Could you please give me some ideas?

Best Regards,
Alison Wang

Quoted 110 lines

-----Original Message-----
From: Alison Wang
Sent: Thursday, December 12, 2019 5:25 PM
To: Oleksandr Suvorov <[email protected]>
Cc: Marcel Ziswiler <[email protected]>; Igor Opaniuk <[email protected]>; [email protected]; [email protected]; [email protected]; [email protected]; [email protected]
Subject: RE: [EXT] Re: [PATCH] ASoC: sgtl5000: Revert "ASoC: sgtl5000: Fix of unmute outputs on probe"

Hi, Oleksandr,

I think I have explained the reason in my email which sent to you yesterday. I attached it too.
According to my test on the boards, MUTE_LO, MUTE_HP and MUTE_ADC are all muted when playing the sound.

Could you give me some ideas about this issue?

Best Regards,I
Alison Wang

Quoted 7 lines

-----Original Message-----
From: Oleksandr Suvorov <[email protected]>
Sent: Thursday, December 12, 2019 5:11 PM
To: Alison Wang <[email protected]>
Cc: Marcel Ziswiler <[email protected]>; Igor Opaniuk <[email protected]>; [email protected]; [email protected]; [email protected]; Oleksandr Suvorov

<[email protected]>;

Quoted 7 lines

[email protected]; [email protected]
Subject: [EXT] Re: [PATCH] ASoC: sgtl5000: Revert "ASoC: sgtl5000: Fix of unmute outputs on probe"

Caution: EXT Email

Alison, could you please explain, what reason to revert this fix? All these

blocks

Quoted 4 lines

have their own control and unmute on demand.
Why do you stand on unconditional unmuting of outputs and ADC on driver probing?

Quoted 8 lines
Quoted 6 lines · Alison Wang

This reverts commit 631bc8f0134ae9620d86a96b8c5f9445d91a2dca.

In commit 631bc8f0134a, snd_soc_component_update_bits is used instead of snd_soc_component_write. Although EN_HP_ZCD and EN_ADC_ZCD are enabled in ANA_CTRL register, MUTE_LO, MUTE_HP and MUTE_ADC bits

are

Quoted 56 lines

sound/soc/codecs/sgtl5000.c static int sgtl5000_probe(struct

Quoted 16 lines

remained as the default value. It causes LO, HP and ADC are all muted after driver probe.

The patch is to revert this commit, snd_soc_component_write is still used and MUTE_LO, MUTE_HP and MUTE_ADC are set as zero (unmuted).

---
 sound/soc/codecs/sgtl5000.c | 6 +++---
 1 file changed, 3 insertions(+), 3 deletions(-)
sound/soc/codecs/sgtl5000.c +0 -0
index aa1f963..0b35fca 100644
@@ -1458,7 +1458,6 @@ static int sgtl5000_probe(struct

snd_soc_component *component)

sound/soc/codecs/sgtl5000.c static int sgtl5000_probe(struct snd_soc_component *component)

Quoted 3 lines
sound/soc/codecs/sgtl5000.c +0 -0
        int ret;
        u16 reg;
        struct sgtl5000_priv *sgtl5000 =

snd_soc_component_get_drvdata(component);

sound/soc/codecs/sgtl5000.c static int sgtl5000_probe(struct · line 1461

Quoted 1 line
sound/soc/codecs/sgtl5000.c +0 -1
-       unsigned int zcd_mask = SGTL5000_HP_ZCD_EN |

SGTL5000_ADC_ZCD_EN;

sound/soc/codecs/sgtl5000.c static int sgtl5000_probe(struct

Quoted 4 lines
sound/soc/codecs/sgtl5000.c +0 -0
        /* power up sgtl5000 */
        ret = sgtl5000_set_power_regs(component);
@@ -1486,8 +1485,9 @@ static int sgtl5000_probe(struct

snd_soc_component *component)

sound/soc/codecs/sgtl5000.c static int sgtl5000_probe(struct · line 1486

Quoted 2 lines
sound/soc/codecs/sgtl5000.c +0 -0
               0x1f);
        snd_soc_component_write(component,

SGTL5000_CHIP_PAD_STRENGTH,

sound/soc/codecs/sgtl5000.c static int sgtl5000_probe(struct · line 1490

Quoted 3 lines
sound/soc/codecs/sgtl5000.c +0 -1
 reg);
 
-       snd_soc_component_update_bits(component,

SGTL5000_CHIP_ANA_CTRL,

sound/soc/codecs/sgtl5000.c static int sgtl5000_probe(struct · line 1489

Quoted 2 lines
sound/soc/codecs/sgtl5000.c +1 -1
-               zcd_mask, zcd_mask);
+       snd_soc_component_write(component,

SGTL5000_CHIP_ANA_CTRL,

sound/soc/codecs/sgtl5000.c static int sgtl5000_probe(struct snd_soc_component *component)

Quoted 4 lines
sound/soc/codecs/sgtl5000.c +2 -0
+                       SGTL5000_HP_ZCD_EN |
+                       SGTL5000_ADC_ZCD_EN);
 
        snd_soc_component_update_bits(component,

SGTL5000_CHIP_MIC_CTRL,

Quoted 4 lines
                        SGTL5000_BIAS_R_MASK,

--
2.9.5

--
Best regards
Oleksandr Suvorov

Toradex AG
Altsagenstrasse 5 | 6048 Horw/Luzern | Switzerland | T: +41 41 500 4800 (main line)

Signature · 6 lines
Best regards
Oleksandr Suvorov

Toradex AG
Altsagenstrasse 5 | 6048 Horw/Luzern | Switzerland | T: +41 41 500
4800 (main line)
#5 lore raw

Hi, Oleksandr,
...
They unmute with standard sound controls:
static const struct snd_kcontrol_new sgtl5000_snd_controls[] = {
...
SOC_SINGLE("Capture Switch", SGTL5000_CHIP_ANA_CTRL, 0, 1, 1),
...
SOC_SINGLE("Headphone Playback Switch", SGTL5000_CHIP_ANA_CTRL, 4, 1, 1),
...
SOC_SINGLE("Lineout Playback Switch", SGTL5000_CHIP_ANA_CTRL, 8, 1, 1),
...

We tested this standard solution using gstreamer and standard ALSA
tools like aplay, arecord and all these tools unmute needed blocks
successfully.
[Alison Wang] I am using aplay. Do you mean I need to add some parameters for aplay or others to unmute the outputs?

Best Regards,
Alison Wang

Quoted 122 lines · Alison Wang

Hi, Oleksandr,

In your patch, outputs and ADC are muted on driver probing. I don't know when and where they can be unmuted in the kernel before playing in the application.

Could you please give me some ideas?

Best Regards,
Alison Wang

Quoted 110 lines

-----Original Message-----
From: Alison Wang
Sent: Thursday, December 12, 2019 5:25 PM
To: Oleksandr Suvorov <[email protected]>
Cc: Marcel Ziswiler <[email protected]>; Igor Opaniuk <[email protected]>; [email protected]; [email protected]; [email protected]; [email protected]; [email protected]
Subject: RE: [EXT] Re: [PATCH] ASoC: sgtl5000: Revert "ASoC: sgtl5000: Fix of unmute outputs on probe"

Hi, Oleksandr,

I think I have explained the reason in my email which sent to you yesterday. I attached it too.
According to my test on the boards, MUTE_LO, MUTE_HP and MUTE_ADC are all muted when playing the sound.

Could you give me some ideas about this issue?

Best Regards,I
Alison Wang

Quoted 7 lines

-----Original Message-----
From: Oleksandr Suvorov <[email protected]>
Sent: Thursday, December 12, 2019 5:11 PM
To: Alison Wang <[email protected]>
Cc: Marcel Ziswiler <[email protected]>; Igor Opaniuk <[email protected]>; [email protected]; [email protected]; [email protected]; Oleksandr Suvorov

<[email protected]>;

Quoted 7 lines

[email protected]; [email protected]
Subject: [EXT] Re: [PATCH] ASoC: sgtl5000: Revert "ASoC: sgtl5000: Fix of unmute outputs on probe"

Caution: EXT Email

Alison, could you please explain, what reason to revert this fix? All these

blocks

Quoted 4 lines

have their own control and unmute on demand.
Why do you stand on unconditional unmuting of outputs and ADC on driver probing?

Quoted 8 lines
Quoted 6 lines · Alison Wang

This reverts commit 631bc8f0134ae9620d86a96b8c5f9445d91a2dca.

In commit 631bc8f0134a, snd_soc_component_update_bits is used instead of snd_soc_component_write. Although EN_HP_ZCD and EN_ADC_ZCD are enabled in ANA_CTRL register, MUTE_LO, MUTE_HP and MUTE_ADC bits

are

Quoted 56 lines

sound/soc/codecs/sgtl5000.c static int sgtl5000_probe(struct

Quoted 16 lines

remained as the default value. It causes LO, HP and ADC are all muted after driver probe.

The patch is to revert this commit, snd_soc_component_write is still used and MUTE_LO, MUTE_HP and MUTE_ADC are set as zero (unmuted).

---
 sound/soc/codecs/sgtl5000.c | 6 +++---
 1 file changed, 3 insertions(+), 3 deletions(-)
sound/soc/codecs/sgtl5000.c +0 -0
index aa1f963..0b35fca 100644
@@ -1458,7 +1458,6 @@ static int sgtl5000_probe(struct

snd_soc_component *component)

sound/soc/codecs/sgtl5000.c static int sgtl5000_probe(struct snd_soc_component *component)

Quoted 3 lines
sound/soc/codecs/sgtl5000.c +0 -0
        int ret;
        u16 reg;
        struct sgtl5000_priv *sgtl5000 =

snd_soc_component_get_drvdata(component);

sound/soc/codecs/sgtl5000.c static int sgtl5000_probe(struct · line 1461

Quoted 1 line
sound/soc/codecs/sgtl5000.c +0 -1
-       unsigned int zcd_mask = SGTL5000_HP_ZCD_EN |

SGTL5000_ADC_ZCD_EN;

sound/soc/codecs/sgtl5000.c static int sgtl5000_probe(struct

Quoted 4 lines
sound/soc/codecs/sgtl5000.c +0 -0
        /* power up sgtl5000 */
        ret = sgtl5000_set_power_regs(component);
@@ -1486,8 +1485,9 @@ static int sgtl5000_probe(struct

snd_soc_component *component)

sound/soc/codecs/sgtl5000.c static int sgtl5000_probe(struct · line 1486

Quoted 2 lines
sound/soc/codecs/sgtl5000.c +0 -0
               0x1f);
        snd_soc_component_write(component,

SGTL5000_CHIP_PAD_STRENGTH,

sound/soc/codecs/sgtl5000.c static int sgtl5000_probe(struct · line 1490

Quoted 3 lines
sound/soc/codecs/sgtl5000.c +0 -1
 reg);
 
-       snd_soc_component_update_bits(component,

SGTL5000_CHIP_ANA_CTRL,

sound/soc/codecs/sgtl5000.c static int sgtl5000_probe(struct · line 1489

Quoted 2 lines
sound/soc/codecs/sgtl5000.c +1 -1
-               zcd_mask, zcd_mask);
+       snd_soc_component_write(component,

SGTL5000_CHIP_ANA_CTRL,

sound/soc/codecs/sgtl5000.c static int sgtl5000_probe(struct snd_soc_component *component)

Quoted 4 lines
sound/soc/codecs/sgtl5000.c +2 -0
+                       SGTL5000_HP_ZCD_EN |
+                       SGTL5000_ADC_ZCD_EN);
 
        snd_soc_component_update_bits(component,

SGTL5000_CHIP_MIC_CTRL,

Quoted 4 lines
                        SGTL5000_BIAS_R_MASK,

--
2.9.5

--
Best regards
Oleksandr Suvorov

Toradex AG
Altsagenstrasse 5 | 6048 Horw/Luzern | Switzerland | T: +41 41 500 4800 (main line)

--
Best regards
Oleksandr Suvorov

Toradex AG
Altsagenstrasse 5 | 6048 Horw/Luzern | Switzerland | T: +41 41 500 4800 (main line)
_______________________________________________
Alsa-devel mailing list
[email protected]
https://mailman.alsa-project.org/mailman/listinfo/alsa-devel

maintainer
#6 lore raw
Quoted 5 lines · Alison Wang

We tested this standard solution using gstreamer and standard ALSA tools like aplay, arecord and all these tools unmute needed blocks successfully.

[Alison Wang] I am using aplay. Do you mean I need to add some parameters for aplay or others to unmute the outputs?

Use amixer.

#7 lore raw
Quoted 5 lines · Alison Wang

We tested this standard solution using gstreamer and standard ALSA tools like aplay, arecord and all these tools unmute needed blocks successfully.

[Alison Wang] I am using aplay. Do you mean I need to add some parameters for aplay or others to unmute the outputs?

Use amixer.
[Alison Wang] Got it. Thanks.

Best Regards,
Alison Wang
_______________________________________________
Alsa-devel mailing list
[email protected]
https://mailman.alsa-project.org/mailman/listinfo/alsa-devel

#8 lore raw
In reply to Mark Brown

On Thu, 2019-12-12 at 12:23 +0000, Mark Brown wrote:

Quoted 5 lines · Alison Wang

We tested this standard solution using gstreamer and standard ALSA tools like aplay, arecord and all these tools unmute needed blocks successfully.

[Alison Wang] I am using aplay. Do you mean I need to add some parameters for aplay or others to unmute the outputs?

Hi Alison,

You can also update your asound.state file with default state for your controls.
_______________________________________________
Alsa-devel mailing list
[email protected]
https://mailman.alsa-project.org/mailman/listinfo/alsa-devel

#9 lore raw

Hi, Daniel,

On Thu, 2019-12-12 at 12:23 +0000, Mark Brown wrote:

Quoted 5 lines · Alison Wang

We tested this standard solution using gstreamer and standard ALSA tools like aplay, arecord and all these tools unmute needed blocks successfully.

[Alison Wang] I am using aplay. Do you mean I need to add some parameters for aplay or others to unmute the outputs?

Hi Alison,

You can also update your asound.state file with default state for your controls. [Alison] Got it. Thanks.

Best Regards,
Alison Wang
_______________________________________________
Alsa-devel mailing list
[email protected]
https://mailman.alsa-project.org/mailman/listinfo/alsa-devel

#10 lore raw
In reply to Mark Brown
Quoted 10 lines · Mark Brown
Quoted 5 lines · Alison Wang

We tested this standard solution using gstreamer and standard ALSA tools like aplay, arecord and all these tools unmute needed blocks successfully.

[Alison Wang] I am using aplay. Do you mean I need to add some parameters for aplay or others to unmute the outputs?

Use amixer.

Marc / Oleksandr,

I can't seem to find the original patch in my mailbox for 631bc8f:
('ASoC: sgtl5000: Fix of unmute outputs on probe') however I find it breaks sgtl5000 audio output on the Gateworks boards which is still broken on 5.6-rc6. Was there some follow-up patches that haven't made it into mainline yet regarding this?

The response above indicates maybe there was an additional ALSA control perhaps added as a resolution but I don't see any differences there.

Best Regards,

Tim

#11 lore raw

Hi Tim

Quoted 12 lines · Tim Harvey
Quoted 10 lines · Mark Brown
Quoted 7 lines · Alison Wang

We tested this standard solution using gstreamer and standard ALSA
tools like aplay, arecord and all these tools unmute needed blocks
successfully.
[Alison Wang] I am using aplay. Do you mean I need to add some parameters for aplay or others to unmute the outputs?

Quoted 6 lines
Quoted 1 line

Use amixer.

Marc / Oleksandr,

I can't seem to find the original patch in my mailbox for 631bc8f:
('ASoC: sgtl5000: Fix of unmute outputs on probe')

I forwarded you that one again. OK?

Quoted 3 lines

however I find it
breaks sgtl5000 audio output on the Gateworks boards which is still broken on 5.6-rc6.

What exactly do you mean by "breaks"? Isn't it that you just need to unmute stuff e.g. using amixer or using a proper updated asound.state file with default states for your controls?

Quoted 2 lines

Was there some follow-up patches that haven't made it into mainline yet regarding this?

I don't think so. It all works perfectly, not?

Quoted 3 lines

The response above indicates maybe there was an additional ALSA control perhaps added as a resolution but I don't see any differences there.

Not that I am aware of, no.

Quoted 3 lines

Best Regards,

Tim

Cheers

Marcel

#12 lore raw
Quoted 25 lines · Marcel Ziswiler

Hi Tim

Quoted 12 lines · Tim Harvey
Quoted 10 lines · Mark Brown
Quoted 7 lines · Alison Wang

We tested this standard solution using gstreamer and standard ALSA
tools like aplay, arecord and all these tools unmute needed blocks
successfully.
[Alison Wang] I am using aplay. Do you mean I need to add some parameters for aplay or others to unmute the outputs?

Quoted 6 lines
Quoted 1 line

Use amixer.

Marc / Oleksandr,

I can't seem to find the original patch in my mailbox for 631bc8f:
('ASoC: sgtl5000: Fix of unmute outputs on probe')

I forwarded you that one again. OK?

Quoted 7 lines
Quoted 3 lines

however I find it
breaks sgtl5000 audio output on the Gateworks boards which is still broken on 5.6-rc6.

What exactly do you mean by "breaks"? Isn't it that you just need to unmute stuff e.g. using amixer or using a proper updated asound.state file with default states for your controls?

the audio device is in /proc/asound/cards but when I send audio to it I 'hear' nothing out the normal line-out output.

Quoted 4 lines
Quoted 3 lines

Was there some follow-up patches that haven't made it into mainline yet regarding this?

Quoted 8 lines

I don't think so. It all works perfectly, not?

Quoted 3 lines

The response above indicates maybe there was an additional ALSA control perhaps added as a resolution but I don't see any differences there.

Not that I am aware of, no.

The output of 'amixer' shows nothing different than before this patch where audio out worked (same controls, same settings on them). I'm testing this with a buildroot rootfs with no asound.conf (or at least none that I know of... i'm honestly not clear where all they can be).

Tim

#13 lore raw
Quoted 49 lines · Tim Harvey
Quoted 25 lines · Marcel Ziswiler

Hi Tim

Quoted 12 lines · Tim Harvey
Quoted 10 lines · Mark Brown
Quoted 7 lines · Alison Wang

We tested this standard solution using gstreamer and standard ALSA
tools like aplay, arecord and all these tools unmute needed blocks
successfully.
[Alison Wang] I am using aplay. Do you mean I need to add some parameters for aplay or others to unmute the outputs?

Quoted 6 lines
Quoted 1 line

Use amixer.

Marc / Oleksandr,

I can't seem to find the original patch in my mailbox for 631bc8f:
('ASoC: sgtl5000: Fix of unmute outputs on probe')

I forwarded you that one again. OK?

Quoted 7 lines
Quoted 3 lines

however I find it
breaks sgtl5000 audio output on the Gateworks boards which is still broken on 5.6-rc6.

What exactly do you mean by "breaks"? Isn't it that you just need to unmute stuff e.g. using amixer or using a proper updated asound.state file with default states for your controls?

the audio device is in /proc/asound/cards but when I send audio to it I 'hear' nothing out the normal line-out output.

Quoted 10 lines
Quoted 3 lines

Was there some follow-up patches that haven't made it into mainline yet regarding this?

I don't think so. It all works perfectly, not?

Quoted 3 lines

The response above indicates maybe there was an additional ALSA control perhaps added as a resolution but I don't see any differences there.

Quoted 7 lines
Quoted 2 lines

Not that I am aware of, no.

The output of 'amixer' shows nothing different than before this patch where audio out worked (same controls, same settings on them). I'm testing this with a buildroot rootfs with no asound.conf (or at least none that I know of... i'm honestly not clear where all they can be).

Tim, did you try to unmute the output with amixer?

Could you provide the output of your amixer with and without this patch?

Before this patch, the driver unmuted HP, LO, and ADC unconditionally on load (while it just had to set up ZCD bits).
Now HP, LO, ADC remain muted until one unmutes them using standard ALSA tools/interfaces.
ALSA mute/unmute controls for these outputs have been presenting in the kernel for a long time. Please, just use them.

Quoted 2 lines

Tim

--
Best regards
Oleksandr Suvorov

Toradex AG
Ebenaustrasse 10 | 6048 Horw | Switzerland | T: +41 41 500 48 00

#14 lore raw
Quoted 61 lines · Oleksandr Suvorov
Quoted 49 lines · Tim Harvey
Quoted 25 lines · Marcel Ziswiler

Hi Tim

Quoted 12 lines · Tim Harvey
Quoted 10 lines · Mark Brown
Quoted 7 lines · Alison Wang

We tested this standard solution using gstreamer and standard ALSA
tools like aplay, arecord and all these tools unmute needed blocks
successfully.
[Alison Wang] I am using aplay. Do you mean I need to add some parameters for aplay or others to unmute the outputs?

Quoted 6 lines
Quoted 1 line

Use amixer.

Marc / Oleksandr,

I can't seem to find the original patch in my mailbox for 631bc8f:
('ASoC: sgtl5000: Fix of unmute outputs on probe')

I forwarded you that one again. OK?

Quoted 7 lines
Quoted 3 lines

however I find it
breaks sgtl5000 audio output on the Gateworks boards which is still broken on 5.6-rc6.

What exactly do you mean by "breaks"? Isn't it that you just need to unmute stuff e.g. using amixer or using a proper updated asound.state file with default states for your controls?

the audio device is in /proc/asound/cards but when I send audio to it I 'hear' nothing out the normal line-out output.

Quoted 10 lines
Quoted 3 lines

Was there some follow-up patches that haven't made it into mainline yet regarding this?

I don't think so. It all works perfectly, not?

Quoted 3 lines

The response above indicates maybe there was an additional ALSA control perhaps added as a resolution but I don't see any differences there.

Quoted 7 lines
Quoted 2 lines

Not that I am aware of, no.

The output of 'amixer' shows nothing different than before this patch where audio out worked (same controls, same settings on them). I'm testing this with a buildroot rootfs with no asound.conf (or at least none that I know of... i'm honestly not clear where all they can be).

Tim, did you try to unmute the output with amixer?

Quoted 9 lines

Could you provide the output of your amixer with and without this patch?

Before this patch, the driver unmuted HP, LO, and ADC unconditionally on load (while it just had to set up ZCD bits).
Now HP, LO, ADC remain muted until one unmutes them using standard ALSA tools/interfaces.
ALSA mute/unmute controls for these outputs have been presenting in the kernel for a long time. Please, just use them.

Oleksandr,

When I first bisected to this I must have done something wrong as I thought amixer settings showed the same before and after - I see that I'm wrong about that. I see the differences now with HP, LO, and ADC muted by default. I agree using amixer controls is fine.

Sorry for the noise!

Tim

maintainer
#15 lore raw
In reply to Tim Harvey
Quoted 3 lines · Tim Harvey

The response above indicates maybe there was an additional ALSA control perhaps added as a resolution but I don't see any differences there.

The response is talking about existing controls that are already in the driver.