[alsa-devel] [RFC 02/15] ASoC: rt5677: keep analog power register at SND_SOC_BIAS_OFF

[Curtis Malainey]

From: Ben Zhang <benzh at chromium.org>

Instead of clearing RT5677_PWR_ANLG2 (MX-64h) to 0 at SND_SOC_BIAS_OFF,
we only clear the RT5677_PWR_CORE bit which is set at SND_SOC_BIAS_PREPARE.
MICBIAS control bits are left unchanged.

This fixed the bug where if MICBIAS1 widget is forced on, MICBIAS
control bits will be cleared at suspend and never turned back on again,
since DAPM thinks the widget is always on.

Signed-off-by: Ben Zhang <benzh at chromium.org>
Signed-off-by: Curtis Malainey <cujomalainey at chromium.org>
---
 sound/soc/codecs/rt5677.c | 4 ++--
 1 file changed, 2 insertions(+), 2 deletions(-)

diff --git a/sound/soc/codecs/rt5677.c b/sound/soc/codecs/rt5677.c
index 5b6ca3ced13b..315a3d39bc09 100644
--- a/sound/soc/codecs/rt5677.c
+++ b/sound/soc/codecs/rt5677.c
@@ -4493,11 +4493,11 @@ static int rt5677_set_bias_level(struct snd_soc_component *component,
 	case SND_SOC_BIAS_OFF:
 		regmap_update_bits(rt5677->regmap, RT5677_DIG_MISC, 0x1, 0x0);
 		regmap_write(rt5677->regmap, RT5677_PWR_DIG1, 0x0000);
-		regmap_write(rt5677->regmap, RT5677_PWR_DIG2, 0x0000);
 		regmap_write(rt5677->regmap, RT5677_PWR_ANLG1,
 			2 << RT5677_LDO1_SEL_SFT |
 			2 << RT5677_LDO2_SEL_SFT);
-		regmap_write(rt5677->regmap, RT5677_PWR_ANLG2, 0x0000);
+		regmap_update_bits(rt5677->regmap, RT5677_PWR_ANLG2,
+			RT5677_PWR_CORE, 0);
 		regmap_update_bits(rt5677->regmap,
 			RT5677_PR_BASE + RT5677_BIAS_CUR4, 0x0f00, 0x0000);
 
-- 
2.23.0.187.g17f5b7556c-goog