ALSA: hda/realtek: Restore headset mic after suspend for Acer Nitro 5
PATCH 1/1 ALSA: hda/realtek: Restore headset mic after suspend for Acer Nitro 5
The headset mic on the AN515-57 was shown as unplugged after resuming from S3/S4, requiring a reboot to recover.
Signed-off-by Breno Baptista <[email protected]>
--- sound/hda/codecs/realtek/alc269.c | 10 ++++++++++ 1 file changed, 10 insertions(+)
sound/hda/codecs/realtek/alc269.c +10 -0
@@ -3578,6 +3578,16 @@ static void alc2xx_fixup_headset_mic(struct hda_codec *codec,alc_update_coef_idx (codec,0x45 ,0xf <<12 |1 <<10 ,5 <<12 ); spec->parse_flags |=HDA_PINCFG_HEADSET_MIC ;break ; +case HDA_FIXUP_ACT_INIT : +/* +* COEF 0x45 (CTIA combo jack mode) is not preserved across suspend. +* Restore it before restarting HP JD so the headset is re-detected. +*/ +if (is_s3_resume (codec) ||is_s4_resume (codec)) { +alc_update_coef_idx (codec,0x45 ,0xf <<12 |1 <<10 ,5 <<12 ); +alc_combo_jack_hp_jd_restart (codec); + } +break ; } }
Signature · 1 line
2.54.0
Quoted 19 lines · Breno Baptista
The headset mic on the AN515-57 was shown as unplugged after resuming from S3/S4, requiring a reboot to recover.
Signed-off-by Breno Baptista <[email protected]>--- sound/hda/codecs/realtek/alc269.c | 10 ++++++++++ 1 file changed, 10 insertions(+)sound/hda/codecs/realtek/alc269.c +2
-0@@ -3578,6 +3578,16 @@ static void alc2xx_fixup_headset_mic(struct hda_codec *codec,alc_update_coef_idx (codec,0x45 ,0xf <<12 |1 <<10 ,5 <<12 ); spec->parse_flags |=HDA_PINCFG_HEADSET_MIC ;break ; +case HDA_FIXUP_ACT_INIT : +/*
sound/hda/codecs/realtek/alc269.c static void alc2xx_fixup_headset_mic(struct hda_codec *codec, · line 3592
Quoted 10 lines
sound/hda/codecs/realtek/alc269.c +8
-0+* COEF 0x45 (CTIA combo jack mode) is not preserved across suspend. +* Restore it before restarting HP JD so the headset is re-detected. +*/ +if (is_s3_resume (codec) ||is_s4_resume (codec)) { +alc_update_coef_idx (codec,0x45 ,0xf <<12 |1 <<10 ,5 <<12 ); +alc_combo_jack_hp_jd_restart (codec); + } +break ; } }
The code change itself looks OK, but I'd like to confirm that it's the expected behavior.
Kailang, could you verify this please?
thanks,
Takashi