ASoc: tas2781: fixed the issue that the chip do not shutdown immediatly after aplay stopped
Subsystems (3)
Shenghao Ding
maintainerPATCH 1/1 ASoc: tas2781: fixed the issue that the chip do not shutdown immediatly after aplay stopped
Issue reported from customer that the chip do not shutdown after aplay stopped until 6 mins later. Drop tasdevice_dapm_event and implement .stream_mute in the tasdevice_dai_ops.
---
v1:
- Drop tasdevice_dapm_event and implement .stream_mute in the tasdevice_dai_ops
- Remove unnecessary line feed for tasdevice_codec_remove
--- sound/soc/codecs/tas2781-i2c.c | 23 ++++++++--------------- 1 file changed, 8 insertions(+), 15 deletions(-)
sound/soc/codecs/tas2781-i2c.c +8 -15
@@ -557,30 +557,22 @@ static void tasdevice_fw_ready(const struct firmware *fmw,release_firmware (fmw); } -static int tasdevice_dapm_event (struct snd_soc_dapm_widget *w, -struct snd_kcontrol *kcontrol,int event) +static int tasdevice_mute (struct snd_soc_dai *dai,int mute,int stream) { -struct snd_soc_component *codec =snd_soc_dapm_to_component (w->dapm); +struct snd_soc_component *codec = dai->component;struct tasdevice_priv *tas_priv =snd_soc_component_get_drvdata (codec); -int state =0 ; -/* Codec Lock Hold */ -mutex_lock (&tas_priv->codec_lock); -if (event ==SND_SOC_DAPM_PRE_PMD ) - state =1 ; -tasdevice_tuning_switch (tas_priv, state); -/* Codec Lock Release*/ -mutex_unlock (&tas_priv->codec_lock); +/* Codec Lock/UnLock */ +guard (mutex)(&tas_priv->codec_lock); +tasdevice_tuning_switch (tas_priv, mute);return 0 ; }static const struct snd_soc_dapm_widget tasdevice_dapm_widgets[] = {SND_SOC_DAPM_AIF_IN ("ASI" ,"ASI Playback" ,0 ,SND_SOC_NOPM ,0 ,0 ), -SND_SOC_DAPM_AIF_OUT_E ("ASI OUT" ,"ASI Capture" ,0 ,SND_SOC_NOPM , -0 ,0 , tasdevice_dapm_event, -SND_SOC_DAPM_POST_PMU |SND_SOC_DAPM_PRE_PMD ), -SND_SOC_DAPM_SPK ("SPK" , tasdevice_dapm_event), +SND_SOC_DAPM_AIF_OUT ("ASI OUT" ,"ASI Capture" ,0 ,SND_SOC_NOPM ,0 ,0 ), +SND_SOC_DAPM_SPK ("SPK" ,NULL ),SND_SOC_DAPM_OUTPUT ("OUT" ),SND_SOC_DAPM_INPUT ("DMIC" ) };
@@ -667,6 +659,7 @@ static const struct snd_soc_dai_ops tasdevice_dai_ops = { .startup = tasdevice_startup, .hw_params = tasdevice_hw_params, .set_sysclk = tasdevice_set_dai_sysclk, + .mute_stream = tasdevice_mute, };static struct snd_soc_dai_driver tasdevice_dai_driver[] = {
Signature · 1 line
2.34.1
Mark Brown
maintainerQuoted 3 lines · Shenghao Ding
Issue reported from customer that the chip do not shutdown after aplay stopped until 6 mins later. Drop
tasdevice_dapm_eventand implement .stream_mute in thetasdevice_dai_ops.
Six minutes sounds like a usersrpace issue with userspace sitting playing silence for a long time rather than a driver issue. By default DAPM does defer powerdown, but only by seconds not minutes.
sound/soc/codecs/tas2781-i2c.c static void tasdevice_fw_ready(const struct firmware *fmw, · line 567
Quoted 5 lines
sound/soc/codecs/tas2781-i2c.c +3
-2-/* Codec Lock Release*/ -mutex_unlock (&tas_priv->codec_lock); +/* Codec Lock/UnLock */ +guard (mutex)(&tas_priv->codec_lock); +tasdevice_tuning_switch (tas_priv, mute);
This is a much heavier weight operation than we're expecting for a mute, it should usually just be literally muting - one or two register writes, not a power up/down sequence
Ding, Shenghao
maintainerHi Broonie
Quoted 29 lines
-----Original Message-----
From: Mark Brown <[email protected]>
Sent: Tuesday, August 20, 2024 12:07 AM
To: Ding, Shenghao <[email protected]>
Cc: [email protected]; [email protected]; [email protected]; [email protected]; [email protected]; [email protected]; [email protected]; Salazar, Ivan <[email protected]>; [email protected]; Yue, Jaden <[email protected]>; [email protected]; Rao, Dipa <[email protected]>; [email protected]; Lo, Henry <[email protected]>; [email protected]; Xu, Baojun <[email protected]>; [email protected]; Chadha, Jasjot Singh <[email protected]>; [email protected]; Navada Kanyana, Mukund <[email protected]>; [email protected]; Kutty, Aanya <[email protected]>; Mahmud, Nayeem <[email protected]>; [email protected]; [email protected]; Ji, Jesse <jesse- [email protected]>; [email protected]; [email protected]; [email protected]; [email protected]
Subject: [EXTERNAL] Re: [PATCH v1] ASoc: tas2781: fixed the issue that the chip do not shutdown immediatly after aplay stoppedQuoted 3 lines · Shenghao Ding
Issue reported from customer that the chip do not shutdown after aplay stopped until 6 mins later. Drop
tasdevice_dapm_eventand implement .stream_mute in thetasdevice_dai_ops.Six minutes sounds like a usersrpace issue with userspace sitting playing silence for a long time rather than a driver issue. By default DAPM does defer powerdown, but only by seconds not minutes.
Quoted 9 lines
sound/soc/codecs/tas2781-i2c.c static void tasdevice_fw_ready(const struct firmware *fmw, · line 567
Quoted 5 lines
sound/soc/codecs/tas2781-i2c.c +3
-2-/* Codec Lock Release*/ -mutex_unlock (&tas_priv->codec_lock); +/* Codec Lock/UnLock */ +guard (mutex)(&tas_priv->codec_lock); +tasdevice_tuning_switch (tas_priv, mute);This is a much heavier weight operation than we're expecting for a mute, it should usually just be literally muting - one or two register writes, not a power up/down sequence
Can the power on is still kept in DAMP, and move the power off(Only set the register 0x2 to 0xe) into the .mute_stream?
Mark Brown
maintainerQuoted 6 lines · Shenghao
Quoted 3 lines
This is a much heavier weight operation than we're expecting for a mute, it should usually just be literally muting - one or two register writes, not a power up/down sequence
Can the power on is still kept in DAMP, and move the power off(Only set the register
0x2to0xe) into the .mute_stream?
Wouldn't that break when a second stream is played, unless whatever is holding the output open is fixed DAPM will still have the device active and therefore not redo the power up sequence?