Skip to contentSkip to conversation

ASoC: Intel: eve: Enable mclk and ssp sclk early

PATCH In review Brent Lu 2 participants 8 messages Oct 1, 2019 – Oct 14, 2019

Subsystems (3)
lore
Share & tools
lore

Brent Lu

author
#0 lore raw

PATCH 1/1 ASoC: Intel: eve: Enable mclk and ssp sclk early

From: Naveen M <[email protected]>

rt5663 and rt5514 needs mclk/sclk early to synchronize its internal clocks.

---
 sound/soc/intel/boards/Kconfig                     |   1 +
 .../soc/intel/boards/kbl_rt5663_rt5514_max98927.c  | 100 +++++++++++++++++++++
 2 files changed, 101 insertions(+)
sound/soc/intel/boards/Kconfig +1 -0
index 5c27f7a..d5f167e 100644
@@ -315,6 +315,7 @@ config SND_SOC_INTEL_KBL_RT5663_RT5514_MAX98927_MACH
 	depends on I2C && ACPI
 	depends on MFD_INTEL_LPSS || COMPILE_TEST
         depends on SPI
+	select SND_SOC_INTEL_SKYLAKE_SSP_CLK
         select SND_SOC_RT5663
         select SND_SOC_RT5514
         select SND_SOC_RT5514_SPI
sound/soc/intel/boards/kbl_rt5663_rt5514_max98927.c +100 -0

147 lines not on this page · raw message

Signature · 6 lines
2.7.4

_______________________________________________
Alsa-devel mailing list
[email protected]
https://mailman.alsa-project.org/mailman/listinfo/alsa-devel
#1 lore raw
Quoted 3 lines

rt5663 and rt5514 needs mclk/sclk early to synchronize its internal clocks.

There are a couple of differences with similar code used in kbl_rt5663_max98927, see below

sound/soc/intel/boards/Kconfig config SND_SOC_INTEL_KBL_RT5663_RT5514_MAX98927_MACH · line 318

Quoted 9 lines
sound/soc/intel/boards/Kconfig +1 -0
index 5c27f7a..d5f167e 100644
@@ -315,6 +315,7 @@ config SND_SOC_INTEL_KBL_RT5663_RT5514_MAX98927_MACH
  	depends on I2C && ACPI
  	depends on MFD_INTEL_LPSS || COMPILE_TEST
          depends on SPI
+	select SND_SOC_INTEL_SKYLAKE_SSP_CLK

It would be nicer to follow the same order as for kbl_rt5663_max98927, with this SKYLAKE_SSP_CLK added last after HDAC_HDMI

Quoted 7 lines
sound/soc/intel/boards/Kconfig +7 -0
+static int platform_clock_control(struct snd_soc_dapm_widget *w,
+			struct snd_kcontrol *k, int  event)
+{
+	struct snd_soc_dapm_context *dapm = w->dapm;
+	struct snd_soc_card *card = dapm->card;
+	struct kbl_codec_private *priv = snd_soc_card_get_drvdata(card);
+	int ret = 0;

sound/soc/intel/boards/Kconfig config SND_SOC_INTEL_KBL_RT5663_RT5514_MAX98927_MACH

Quoted 10 lines
sound/soc/intel/boards/Kconfig +10 -0
+
+	/*
+	 * MCLK/SCLK need to be ON early for a successful synchronization of
+	 * codec internal clock. And the clocks are turned off during
+	 * POST_PMD after the stream is stopped.
+	 */
+	switch (event) {
+	case SND_SOC_DAPM_PRE_PMU:
+		if (__clk_is_enabled(priv->mclk))
+			return 0;

Is this if() test needed? it's not part of the code for kbl_rt5663_max98927, despite all the comments and code structure being identical.

Quoted 23 lines
sound/soc/intel/boards/Kconfig +23 -0
+
+		/* Enable MCLK */
+		ret = clk_set_rate(priv->mclk, 24000000);
+		if (ret < 0) {
+			dev_err(card->dev, "Can't set rate for mclk, err: %d\n",
+				ret);
+			return ret;
+		}
+
+		ret = clk_prepare_enable(priv->mclk);
+		if (ret < 0) {
+			dev_err(card->dev, "Can't enable mclk, err: %d\n", ret);
+			return ret;
+		}
+
+		/* Enable SCLK */
+		ret = clk_set_rate(priv->sclk, 3072000);
+		if (ret < 0) {
+			dev_err(card->dev, "Can't set rate for sclk, err: %d\n",
+				ret);
+			clk_disable_unprepare(priv->mclk);
+			return ret;
+		}

sound/soc/intel/boards/Kconfig

Quoted 10 lines
sound/soc/intel/boards/Kconfig +10 -0
+
+		ret = clk_prepare_enable(priv->sclk);
+		if (ret < 0) {
+			dev_err(card->dev, "Can't enable sclk, err: %d\n", ret);
+			clk_disable_unprepare(priv->mclk);
+		}
+		break;
+	case SND_SOC_DAPM_POST_PMD:
+		if (!__clk_is_enabled(priv->mclk))
+			return 0;

same here, is this if() test needed? If yes, isn't it needed in kbl_rt5663_max98927?

sound/soc/intel/boards/Kconfig

Quoted 11 lines
sound/soc/intel/boards/Kconfig +11 -0
+
+		clk_disable_unprepare(priv->mclk);
+		clk_disable_unprepare(priv->sclk);
+		break;
+	default:
+		return 0;
+	}
+
+	return 0;
+}
+

While I am at it, this machine driver uses .ignore_pmdown_time, which is typically used to avoid clock issues. Now that you have an explicit control on clocks, is .ignore_pmdown_time actually required?
_______________________________________________
Alsa-devel mailing list
[email protected]
https://mailman.alsa-project.org/mailman/listinfo/alsa-devel

author
#2 lore raw
Quoted 7 lines

Review

Quoted 1 line
(quoted diff) +0 -0

b/sound/soc/intel/boards/Kconfig

sound/soc/intel/boards/Kconfig config

Quoted 4 lines

index 5c27f7a..d5f167e 100644

sound/soc/intel/boards/Kconfig +0 -0
@@ -315,6 +315,7 @@ config

SND_SOC_INTEL_KBL_RT5663_RT5514_MAX98927_MACH

Quoted 9 lines

sound/soc/intel/boards/Kconfig config SND_SOC_INTEL_KBL_RT5663_RT5514_MAX98927_MACH · line 318

Quoted 4 lines
sound/soc/intel/boards/Kconfig +1 -0
  	depends on I2C && ACPI
  	depends on MFD_INTEL_LPSS || COMPILE_TEST
          depends on SPI
+	select SND_SOC_INTEL_SKYLAKE_SSP_CLK

It would be nicer to follow the same order as for kbl_rt5663_max98927, with this SKYLAKE_SSP_CLK added last after HDAC_HDMI

Will fix it.

Quoted 13 lines
Quoted 13 lines
sound/soc/intel/boards/Kconfig +13 -0
+static int platform_clock_control(struct snd_soc_dapm_widget *w,
+			struct snd_kcontrol *k, int  event)
+{
+	struct snd_soc_dapm_context *dapm = w->dapm;
+	struct snd_soc_card *card = dapm->card;
+	struct kbl_codec_private *priv = snd_soc_card_get_drvdata(card);
+	int ret = 0;
+
+	/*
+	 * MCLK/SCLK need to be ON early for a successful synchronization of
+	 * codec internal clock. And the clocks are turned off during
+	 * POST_PMD after the stream is stopped.
+	 */
Quoted 9 lines

sound/soc/intel/boards/Kconfig

Quoted 4 lines
sound/soc/intel/boards/Kconfig +4 -0
+	switch (event) {
+	case SND_SOC_DAPM_PRE_PMU:
+		if (__clk_is_enabled(priv->mclk))
+			return 0;

Is this if() test needed? it's not part of the code for kbl_rt5663_max98927, despite all the comments and code structure being identical.

This function call prevents the clk_set_rate() from being called when clock is enabled. It's removed during the upstream review and use the flag CLK_SET_RATE_GATE instead. Will fix it.

https://patchwork.kernel.org/patch/10070383/
https://patchwork.kernel.org/patch/10133179/

Quoted 31 lines

sound/soc/intel/boards/Kconfig

Quoted 5 lines
sound/soc/intel/boards/Kconfig +5 -0
+
+		/* Enable MCLK */
+		ret = clk_set_rate(priv->mclk, 24000000);
+		if (ret < 0) {
+			dev_err(card->dev, "Can't set rate for mclk, err:

%d\n",

sound/soc/intel/boards/Kconfig

Quoted 7 lines
sound/soc/intel/boards/Kconfig +7 -0
+				ret);
+			return ret;
+		}
+
+		ret = clk_prepare_enable(priv->mclk);
+		if (ret < 0) {
+			dev_err(card->dev, "Can't enable mclk, err: %d\n",

ret);

sound/soc/intel/boards/Kconfig

Quoted 7 lines
sound/soc/intel/boards/Kconfig +7 -0
+			return ret;
+		}
+
+		/* Enable SCLK */
+		ret = clk_set_rate(priv->sclk, 3072000);
+		if (ret < 0) {
+			dev_err(card->dev, "Can't set rate for sclk, err:

%d\n",

sound/soc/intel/boards/Kconfig

Quoted 8 lines
sound/soc/intel/boards/Kconfig +8 -0
+				ret);
+			clk_disable_unprepare(priv->mclk);
+			return ret;
+		}
+
+		ret = clk_prepare_enable(priv->sclk);
+		if (ret < 0) {
+			dev_err(card->dev, "Can't enable sclk, err: %d\n",

ret);

Quoted 10 lines

sound/soc/intel/boards/Kconfig

Quoted 6 lines
sound/soc/intel/boards/Kconfig +6 -0
+			clk_disable_unprepare(priv->mclk);
+		}
+		break;
+	case SND_SOC_DAPM_POST_PMD:
+		if (!__clk_is_enabled(priv->mclk))
+			return 0;

same here, is this if() test needed? If yes, isn't it needed in kbl_rt5663_max98927?

Will fix it.

Quoted 7 lines
Quoted 7 lines
sound/soc/intel/boards/Kconfig +7 -0
+
+		clk_disable_unprepare(priv->mclk);
+		clk_disable_unprepare(priv->sclk);
+		break;
+	default:
+		return 0;
+	}
Quoted 8 lines

sound/soc/intel/boards/Kconfig

Quoted 4 lines
sound/soc/intel/boards/Kconfig +4 -0
+
+	return 0;
+}
+

While I am at it, this machine driver uses .ignore_pmdown_time, which is typically used to avoid clock issues. Now that you have an explicit control on clocks, is .ignore_pmdown_time actually required?

Actually I don't know the purpose to use ignore_pmdown_time flag in the first place but I think they are not related since this patch is to turn on ssp clocks earlier while the ignore_pmdown_time flag is used to delay the DAPM power down for playback stream in the soc_pcm_close(). Just my two cents and thank you for the review.

Regards,
Brent

_______________________________________________
Alsa-devel mailing list
[email protected]
https://mailman.alsa-project.org/mailman/listinfo/alsa-devel

author
#3 lore raw

PATCH 1/1 ASoC: Intel: eve: Enable mclk and ssp sclk early

In reply to Brent Lu

From: Naveen M <[email protected]>

rt5663 and rt5514 needs mclk/sclk early to synchronize its internal clocks.

---
 sound/soc/intel/boards/Kconfig                     |  1 +
 .../soc/intel/boards/kbl_rt5663_rt5514_max98927.c  | 94 ++++++++++++++++++++++
 2 files changed, 95 insertions(+)
sound/soc/intel/boards/Kconfig +1 -0
index 5c27f7a..5e0e7db 100644
@@ -320,6 +320,7 @@ config SND_SOC_INTEL_KBL_RT5663_RT5514_MAX98927_MACH
         select SND_SOC_RT5514_SPI
         select SND_SOC_MAX98927
         select SND_SOC_HDAC_HDMI
+	select SND_SOC_INTEL_SKYLAKE_SSP_CLK
         help
           This adds support for ASoC Onboard Codec I2S machine driver. This will
           create an alsa sound card for RT5663 + RT5514 + MAX98927.
sound/soc/intel/boards/kbl_rt5663_rt5514_max98927.c +94 -0

141 lines not on this page · raw message

Signature · 6 lines
2.7.4

_______________________________________________
Alsa-devel mailing list
[email protected]
https://mailman.alsa-project.org/mailman/listinfo/alsa-devel
author
#4 lore raw
Quoted 10 lines

Subject: [PATCH] ASoC: Intel: eve: Enable mclk and ssp sclk early

From: Naveen M <[email protected]>

rt5663 and rt5514 needs mclk/sclk early to synchronize its internal clocks.

Quoted 6 lines
---
 sound/soc/intel/boards/Kconfig                     |  1 +
 .../soc/intel/boards/kbl_rt5663_rt5514_max98927.c  | 94
++++++++++++++++++++++
 2 files changed, 95 insertions(+)

Hi Pierre-Louis

This is the v2 patch based on your previous comment to remove unnecessary function calls. The order in Kconfig is also rearranged. Please let me know if this patch is ok. Thanks.

Quoted 176 lines
(quoted diff) +0 -0

b/sound/soc/intel/boards/Kconfig index 5c27f7a..5e0e7db 100644

sound/soc/intel/boards/Kconfig +0 -0
@@ -320,6 +320,7 @@ config

SND_SOC_INTEL_KBL_RT5663_RT5514_MAX98927_MACH

         select SND_SOC_RT5514_SPI
         select SND_SOC_MAX98927
         select SND_SOC_HDAC_HDMI
sound/soc/intel/boards/Kconfig +1 -0
+	select SND_SOC_INTEL_SKYLAKE_SSP_CLK
         help
           This adds support for ASoC Onboard Codec I2S machine driver. This will
           create an alsa sound card for RT5663 + RT5514 + MAX98927.
sound/soc/intel/boards/kbl_rt5663_rt5514_max98927.c +38 -0
index 74dda87..dc09a85 100644
@@ -22,6 +22,9 @@
 #include "../../codecs/rt5514.h"
 #include "../../codecs/rt5663.h"
 #include "../../codecs/hdac_hdmi.h"
+#include <linux/clk.h>
+#include <linux/clk-provider.h>
+#include <linux/clkdev.h>
 
 #define KBL_REALTEK_CODEC_DAI "rt5663-aif"
 #define KBL_REALTEK_DMIC_CODEC_DAI "rt5514-aif1"
@@ -50,6 +53,8 @@ struct kbl_codec_private {
 	struct snd_soc_jack kabylake_headset;
 	struct list_head hdmi_pcm_list;
 	struct snd_soc_jack kabylake_hdmi[2];
+	struct clk *mclk;
+	struct clk *sclk;
 };
 
 enum {
@@ -71,6 +76,61 @@ static const struct snd_kcontrol_new
 kabylake_controls[] = {
 	SOC_DAPM_PIN_SWITCH("DMIC"),
 };
 
+static int platform_clock_control(struct snd_soc_dapm_widget *w,
+			struct snd_kcontrol *k, int  event)
+{
+	struct snd_soc_dapm_context *dapm = w->dapm;
+	struct snd_soc_card *card = dapm->card;
+	struct kbl_codec_private *priv = snd_soc_card_get_drvdata(card);
+	int ret = 0;
+
+	/*
+	 * MCLK/SCLK need to be ON early for a successful synchronization of
+	 * codec internal clock. And the clocks are turned off during
+	 * POST_PMD after the stream is stopped.
+	 */
+	switch (event) {
+	case SND_SOC_DAPM_PRE_PMU:
+		/* Enable MCLK */
+		ret = clk_set_rate(priv->mclk, 24000000);
+		if (ret < 0) {
+			dev_err(card->dev, "Can't set rate for mclk, err:
 %d\n",
+				ret);
+			return ret;
+		}
+
+		ret = clk_prepare_enable(priv->mclk);
+		if (ret < 0) {
+			dev_err(card->dev, "Can't enable mclk, err: %d\n",
 ret);
+			return ret;
+		}
+
+		/* Enable SCLK */
+		ret = clk_set_rate(priv->sclk, 3072000);
+		if (ret < 0) {
+			dev_err(card->dev, "Can't set rate for sclk, err:

%d\n",

sound/soc/intel/boards/kbl_rt5663_rt5514_max98927.c +8 -0
+				ret);
+			clk_disable_unprepare(priv->mclk);
+			return ret;
+		}
+
+		ret = clk_prepare_enable(priv->sclk);
+		if (ret < 0) {
+			dev_err(card->dev, "Can't enable sclk, err: %d\n",

ret);

sound/soc/intel/boards/kbl_rt5663_rt5514_max98927.c +14 -0
+			clk_disable_unprepare(priv->mclk);
+		}
+		break;
+	case SND_SOC_DAPM_POST_PMD:
+		clk_disable_unprepare(priv->mclk);
+		clk_disable_unprepare(priv->sclk);
+		break;
+	default:
+		return 0;
+	}
+
+	return 0;
+}
+
 static const struct snd_soc_dapm_widget kabylake_widgets[] = {
 	SND_SOC_DAPM_HP("Headphone Jack", NULL),
 	SND_SOC_DAPM_MIC("Headset Mic", NULL), @@ -79,11 +139,15

@@ static const struct snd_soc_dapm_widget kabylake_widgets[] = {

sound/soc/intel/boards/kbl_rt5663_rt5514_max98927.c +2 -0
 	SND_SOC_DAPM_MIC("DMIC", NULL),
 	SND_SOC_DAPM_SPK("HDMI1", NULL),
 	SND_SOC_DAPM_SPK("HDMI2", NULL),
+	SND_SOC_DAPM_SUPPLY("Platform Clock", SND_SOC_NOPM, 0, 0,
+			platform_clock_control, SND_SOC_DAPM_PRE_PMU

|

sound/soc/intel/boards/kbl_rt5663_rt5514_max98927.c +32 -0
+			SND_SOC_DAPM_POST_PMD),
 
 };
 
 static const struct snd_soc_dapm_route kabylake_map[] = {
 	/* Headphones */
+	{ "Headphone Jack", NULL, "Platform Clock" },
 	{ "Headphone Jack", NULL, "HPOL" },
 	{ "Headphone Jack", NULL, "HPOR" },
@@ -92,6 +156,7 @@ static const struct snd_soc_dapm_route
 kabylake_map[] = {
 	{ "Right Spk", NULL, "Right BE_OUT" },
 
 	/* other jacks */
+	{ "Headset Mic", NULL, "Platform Clock" },
 	{ "IN1P", NULL, "Headset Mic" },
 	{ "IN1N", NULL, "Headset Mic" },
@@ -643,6 +708,7 @@ static int kabylake_audio_probe(struct
 platform_device *pdev)  {
 	struct kbl_codec_private *ctx;
 	struct snd_soc_acpi_mach *mach;
+	int ret = 0;
 
 	ctx = devm_kzalloc(&pdev->dev, sizeof(*ctx), GFP_KERNEL);
 	if (!ctx)
@@ -658,6 +724,34 @@ static int kabylake_audio_probe(struct
 platform_device *pdev)
 		dmic_constraints = mach->mach_params.dmic_num == 2 ?
 			&constraints_dmic_2ch :
 &constraints_dmic_channels;
 
+	ctx->mclk = devm_clk_get(&pdev->dev, "ssp1_mclk");
+	if (IS_ERR(ctx->mclk)) {
+		ret = PTR_ERR(ctx->mclk);
+		if (ret == -ENOENT) {
+			dev_info(&pdev->dev,
+				"Failed to get ssp1_mclk, defer probe\n");
+			return -EPROBE_DEFER;
+		}
+
+		dev_err(&pdev->dev, "Failed to get ssp1_mclk with
 err:%d\n",
+								ret);
+		return ret;
+	}
+
+	ctx->sclk = devm_clk_get(&pdev->dev, "ssp1_sclk");
+	if (IS_ERR(ctx->sclk)) {
+		ret = PTR_ERR(ctx->sclk);
+		if (ret == -ENOENT) {
+			dev_info(&pdev->dev,
+				"Failed to get ssp1_sclk, defer probe\n");
+			return -EPROBE_DEFER;
+		}
+
+		dev_err(&pdev->dev, "Failed to get ssp1_sclk with err:%d\n",
+								ret);
+		return ret;
+	}
+

return devm_snd_soc_register_card(&pdev->dev,
&kabylake_audio_card); }

Quoted 2 lines

--
2.7.4

_______________________________________________
Alsa-devel mailing list
[email protected]
https://mailman.alsa-project.org/mailman/listinfo/alsa-devel

#5 lore raw
Quoted 17 lines · Brent
Quoted 10 lines

Subject: [PATCH] ASoC: Intel: eve: Enable mclk and ssp sclk early

From: Naveen M <[email protected]>

rt5663 and rt5514 needs mclk/sclk early to synchronize its internal clocks.

Quoted 6 lines
---
  sound/soc/intel/boards/Kconfig                     |  1 +
  .../soc/intel/boards/kbl_rt5663_rt5514_max98927.c  | 94
++++++++++++++++++++++
  2 files changed, 95 insertions(+)
Quoted 5 lines

Hi Pierre-Louis

This is the v2 patch based on your previous comment to remove unnecessary function calls. The order in Kconfig is also rearranged. Please let me know if this patch is ok. Thanks.

I don't see a v2?
_______________________________________________
Alsa-devel mailing list
[email protected]
https://mailman.alsa-project.org/mailman/listinfo/alsa-devel

author
#6 lore raw
Quoted 8 lines
Quoted 6 lines

Hi Pierre-Louis

This is the v2 patch based on your previous comment to remove unnecessary function calls. The order in Kconfig is also rearranged. Please let me know if this patch is ok. Thanks.

I don't see a v2?

Sorry I forgot to use --subject-prefix option when making the patch file. Should I submit a new one with v2 prefix?

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

#7 lore raw
Quoted 11 lines · Brent
Quoted 8 lines
Quoted 6 lines

Hi Pierre-Louis

This is the v2 patch based on your previous comment to remove unnecessary function calls. The order in Kconfig is also rearranged. Please let me know if this patch is ok. Thanks.

I don't see a v2?

Sorry I forgot to use --subject-prefix option when making the patch file. Should I submit a new one with v2 prefix?

yes please, thanks!
_______________________________________________
Alsa-devel mailing list
[email protected]
https://mailman.alsa-project.org/mailman/listinfo/alsa-devel