ASoC: rt1015p: delay 300ms for waiting calibration
Subsystems (3)
v2 cover ASoC: rt1015p: delay 300ms for waiting calibration
The 1st patch moves SDB control from DAI ops trigger to DAPM event (per review comments in v1).
The 2nd patch adds the 300ms delay for waiting calibration.
Changes from v1:
(https://patchwork.kernel.org/project/alsa-devel/patch/[email protected]/)
- Move the delay from trigger to DAPM event.
Tzung-Bi Shih (2):
ASoC: rt1015p: move SDB control from trigger to DAPM
ASoC: rt1015p: delay 300ms after SDB pulling high for calibration
sound/soc/codecs/rt1015p.c | 54 +++++++++++++------------------------- 1 file changed, 18 insertions(+), 36 deletions(-)
Signature · 1 line
2.29.2.576.ga3fc446d84-goog
PATCH v2 1/2 ASoC: rt1015p: move SDB control from trigger to DAPM
Moves SDB control from DAI ops trigger to DAPM.
As long as BCLK and LRCLK are ready, SDB can be toggled earlier.
--- sound/soc/codecs/rt1015p.c | 49 ++++++++++---------------------------- 1 file changed, 12 insertions(+), 37 deletions(-)
sound/soc/codecs/rt1015p.c +12 -37
@@ -19,60 +19,40 @@struct rt1015p_priv {struct gpio_desc *sdb; -int sdb_switch; }; -static int rt1015p_daiops_trigger (struct snd_pcm_substream *substream, -int cmd,struct snd_soc_dai *dai) +static int rt1015p_sdb_event (struct snd_soc_dapm_widget *w, +struct snd_kcontrol *kcontrol,int event) { -struct snd_soc_component *component = dai->component; +struct snd_soc_component *component = +snd_soc_dapm_to_component (w->dapm);struct rt1015p_priv *rt1015p =snd_soc_component_get_drvdata (component);if (!rt1015p->sdb)return 0 ; -switch (cmd) { -case SNDRV_PCM_TRIGGER_START : -case SNDRV_PCM_TRIGGER_RESUME : -case SNDRV_PCM_TRIGGER_PAUSE_RELEASE : -if (rt1015p->sdb_switch) { -gpiod_set_value (rt1015p->sdb,1 ); -dev_dbg (component->dev,"set sdb to 1" ); - } +switch (event) { +case SND_SOC_DAPM_PRE_PMU : +gpiod_set_value (rt1015p->sdb,1 ); +dev_dbg (component->dev,"set sdb to 1" );break ; -case SNDRV_PCM_TRIGGER_STOP : -case SNDRV_PCM_TRIGGER_SUSPEND : -case SNDRV_PCM_TRIGGER_PAUSE_PUSH : +case SND_SOC_DAPM_POST_PMD :gpiod_set_value (rt1015p->sdb,0 );dev_dbg (component->dev,"set sdb to 0" );break ; +default : +break ; }return 0 ; } -static int rt1015p_sdb_event (struct snd_soc_dapm_widget *w, -struct snd_kcontrol *kcontrol,int event) -{ -struct snd_soc_component *component = -snd_soc_dapm_to_component (w->dapm); -struct rt1015p_priv *rt1015p = -snd_soc_component_get_drvdata (component); - -if (event &SND_SOC_DAPM_POST_PMU ) - rt1015p->sdb_switch =1 ; -else if (event &SND_SOC_DAPM_POST_PMD ) - rt1015p->sdb_switch =0 ; - -return 0 ; -} -static const struct snd_soc_dapm_widget rt1015p_dapm_widgets[] = {SND_SOC_DAPM_OUTPUT ("Speaker" ),SND_SOC_DAPM_OUT_DRV_E ("SDB" ,SND_SOC_NOPM ,0 ,0 ,NULL ,0 , rt1015p_sdb_event, -SND_SOC_DAPM_POST_PMU |SND_SOC_DAPM_POST_PMD ), +SND_SOC_DAPM_PRE_PMU |SND_SOC_DAPM_POST_PMD ), };static const struct snd_soc_dapm_route rt1015p_dapm_routes[] = {
@@ -91,10 +71,6 @@ static const struct snd_soc_component_driver rt1015p_component_driver = { .non_legacy_dai_naming = 1, }; -static const struct snd_soc_dai_ops rt1015p_dai_ops = { - .trigger = rt1015p_daiops_trigger, -}; -static struct snd_soc_dai_driver rt1015p_dai_driver = { .name ="HiFi" , .playback = {
@@ -104,7 +80,6 @@ static struct snd_soc_dai_driver rt1015p_dai_driver = { .channels_min = 1, .channels_max = 2, }, - .ops = &rt1015p_dai_ops, };static int rt1015p_platform_probe (struct platform_device *pdev)
Signature · 1 line
2.29.2.576.ga3fc446d84-goog
Mark Brown
maintainerRe: [PATCH v2 1/2] ASoC: rt1015p: move SDB control from trigger to DAPM
sound/soc/codecs/rt1015p.c line 38
Quoted 4 lines · Tzung-Bi Shih
sound/soc/codecs/rt1015p.c +4
-0+switch (event) { +case SND_SOC_DAPM_PRE_PMU : +gpiod_set_value (rt1015p->sdb,1 ); +dev_dbg (component->dev,"set sdb to 1" );
Now this is in DAPM it's not in atomic context so it'd be more friendly to use gpiod_set_value_cansleep() so that it'll work even if the GPIO isn't atomic safe.
Re: [PATCH v2 1/2] ASoC: rt1015p: move SDB control from trigger to DAPM
Quoted 11 lines · Mark Brown
sound/soc/codecs/rt1015p.c line 38
Quoted 4 lines · Tzung-Bi Shih
sound/soc/codecs/rt1015p.c +4
-0+switch (event) { +case SND_SOC_DAPM_PRE_PMU : +gpiod_set_value (rt1015p->sdb,1 ); +dev_dbg (component->dev,"set sdb to 1" );Now this is in DAPM it's not in atomic context so it'd be more friendly to use
gpiod_set_value_cansleep()so that it'll work even if the GPIO isn't atomic safe.
Strange, I thought I should get a warning message because of the following statement in gpiod_set_value (but it didn't):WARN_ON(desc->gdev->chip->can_sleep);
Also I dumped the value from gpiod_cansleep(rt1015p->sdb), it returned 0.
Is it safe to call gpiod_set_value_cansleep() even if the GPIO controller doesn't support it? (I guess yes, _cansleep could be just an advice)
Mark Brown
maintainerRe: [PATCH v2 1/2] ASoC: rt1015p: move SDB control from trigger to DAPM
Quoted 10 lines · Tzung-Bi Shih
Quoted 3 lines · Mark Brown
Now this is in DAPM it's not in atomic context so it'd be more friendly to use
gpiod_set_value_cansleep()so that it'll work even if the GPIO isn't atomic safe.Strange, I thought I should get a warning message because of the following statement in
gpiod_set_value(but it didn't):WARN_ON(desc->gdev->chip->can_sleep);
Quoted 5 lines
Also I dumped the value from
gpiod_cansleep(rt1015p->sdb), it returned 0.Is it safe to call
gpiod_set_value_cansleep()even if the GPIO controller doesn't support it? (I guess yes,_cansleepcould be just an advice)
Yes, absolutely - it's the other way around. _cansleep() is the caller saying that they support using a GPIO that might sleep when updating (eg, one connected over I2C), it works with both GPIOs that might sleep and GPIOs that are atomic safe. Without _cansleep() only atomic safe GPIO controllers can be used.
PATCH v2 2/2 ASoC: rt1015p: delay 300ms after SDB pulling high for calibration
In reply to Tzung-Bi ShihRT1015p needs 300ms delay after SDB pulling high for internal calibration during the power on sequence.
Delays 300ms right before data sends out to avoid data truncated.
--- sound/soc/codecs/rt1015p.c | 7 +++++++ 1 file changed, 7 insertions(+)
sound/soc/codecs/rt1015p.c +7 -0
@@ -4,6 +4,7 @@// // Copyright 2020 The Linux Foundation. All rights reserved. +#include <linux/delay.h> #include <linux/device.h> #include <linux/err.h> #include <linux/gpio.h>
@@ -19,6 +20,7 @@struct rt1015p_priv {struct gpio_desc *sdb; +bool calib_done; }; static int rt1015p_sdb_event(struct snd_soc_dapm_widget *w,
@@ -36,6 +38,11 @@ static int rt1015p_sdb_event(struct snd_soc_dapm_widget *w,case SND_SOC_DAPM_PRE_PMU :gpiod_set_value (rt1015p->sdb,1 );dev_dbg (component->dev,"set sdb to 1" ); + +if (!rt1015p->calib_done) { +msleep (300 ); + rt1015p->calib_done = true; + }break ;case SND_SOC_DAPM_POST_PMD :gpiod_set_value (rt1015p->sdb,0 );
Signature · 1 line
2.29.2.576.ga3fc446d84-goog
Mark Brown
maintainerRe: [PATCH v2 2/2] ASoC: rt1015p: delay 300ms after SDB pulling high for calibration
sound/soc/codecs/rt1015p.c static int rt1015p_sdb_event(struct snd_soc_dapm_widget *w, · line 45
Quoted 4 lines · Tzung-Bi Shih
sound/soc/codecs/rt1015p.c +4
-0+if (!rt1015p->calib_done) { +msleep (300 ); + rt1015p->calib_done = true; + }
Might we need to reset calib_done over suspend? If the device looses power I guess it forgets the calibration.
Re: [PATCH v2 2/2] ASoC: rt1015p: delay 300ms after SDB pulling high for calibration
Quoted 10 lines · Mark Brown
sound/soc/codecs/rt1015p.c static int rt1015p_sdb_event(struct snd_soc_dapm_widget *w, · line 45
Quoted 4 lines · Tzung-Bi Shih
sound/soc/codecs/rt1015p.c +4
-0+if (!rt1015p->calib_done) { +msleep (300 ); + rt1015p->calib_done = true; + }Might we need to reset
calib_doneover suspend? If the device looses power I guess it forgets the calibration.
In our environment, the power VBAT and AVDD are still on even if the system suspends. In more low power mode, AVDD could probably lose.
But agree with you, in general, suppose the device needs to calibrate again after suspend/resume is better.
Mark Brown
maintainerRe: [PATCH v2 2/2] ASoC: rt1015p: delay 300ms after SDB pulling high for calibration
Quoted 10 lines · Tzung-Bi Shih
Quoted 2 lines · Mark Brown
Might we need to reset
calib_doneover suspend? If the device looses power I guess it forgets the calibration.In our environment, the power VBAT and AVDD are still on even if the system suspends. In more low power mode, AVDD could probably lose.
But agree with you, in general, suppose the device needs to calibrate again after suspend/resume is better.
Yeah, even on that system there might be a problem if the system goes into hibernate instead of suspend (especially if it hibernates due to running out of battery!).