From: wangdicheng <wangdicheng@kylinos.cn>

During S3 sleep state, if headphones are unplugged, we need to re-report
their status during the wake phase to ensure proper headphone presence
detection after system resume.

Signed-off-by: wangdicheng <wangdicheng@kylinos.cn>
---

v1->v2:
Problem: No audio output device after S3.
Cause: The audio driver misses re-initialization upon resume,
resulting in stale jack status (remaining in the pre-sleep state),
which prevents the system from finding the correct output device.

 sound/hda/codecs/conexant.c | 8 ++++++++
 1 file changed, 8 insertions(+)

diff --git a/sound/hda/codecs/conexant.c b/sound/hda/codecs/conexant.c
index 5fcbc1312c69..146ff65ccc30 100644
--- a/sound/hda/codecs/conexant.c
+++ b/sound/hda/codecs/conexant.c
@@ -264,6 +264,13 @@ static int cx_suspend(struct hda_codec *codec)
 	return 0;
 }
 
+static int cx_resume(struct hda_codec *codec)
+{
+	cx_init(codec);
+	snd_hda_jack_report_sync(codec);
+	return 0;
+}
+
 /*
  * pin fix-up
  */
@@ -1277,6 +1284,7 @@ static const struct hda_codec_ops cx_codec_ops = {
 	.init = cx_init,
 	.unsol_event = snd_hda_jack_unsol_event,
 	.suspend = cx_suspend,
+	.resume = cx_resume,
 	.check_power_status = snd_hda_gen_check_power_status,
 	.stream_pm = snd_hda_gen_stream_pm,
 };
-- 
2.25.1

On Wed, 19 Nov 2025 09:53:45 +0100,
wangdich9700@163.com wrote:
> 
> From: wangdicheng <wangdicheng@kylinos.cn>
> 
> During S3 sleep state, if headphones are unplugged, we need to re-report
> their status during the wake phase to ensure proper headphone presence
> detection after system resume.
> 
> Signed-off-by: wangdicheng <wangdicheng@kylinos.cn>
> ---
> 
> v1->v2:
> Problem: No audio output device after S3.
> Cause: The audio driver misses re-initialization upon resume,
> resulting in stale jack status (remaining in the pre-sleep state),
> which prevents the system from finding the correct output device.

Again, please double-check what is the really cause.

snd_hda_jack_report_sync() *is* called already in
hda_call_codec_resume().  So snd_hda_jack_report_sync() is just called
twice with your patch, which must be superfluous.

The other significant difference with your patch is that you dropped
snd_hda_regmap_sync().  So it must break something else.


Takashi

> 
>  sound/hda/codecs/conexant.c | 8 ++++++++
>  1 file changed, 8 insertions(+)
> 
> diff --git a/sound/hda/codecs/conexant.c b/sound/hda/codecs/conexant.c
> index 5fcbc1312c69..146ff65ccc30 100644
> --- a/sound/hda/codecs/conexant.c
> +++ b/sound/hda/codecs/conexant.c
> @@ -264,6 +264,13 @@ static int cx_suspend(struct hda_codec *codec)
>  	return 0;
>  }
>  
> +static int cx_resume(struct hda_codec *codec)
> +{
> +	cx_init(codec);
> +	snd_hda_jack_report_sync(codec);
> +	return 0;
> +}
> +
>  /*
>   * pin fix-up
>   */
> @@ -1277,6 +1284,7 @@ static const struct hda_codec_ops cx_codec_ops = {
>  	.init = cx_init,
>  	.unsol_event = snd_hda_jack_unsol_event,
>  	.suspend = cx_suspend,
> +	.resume = cx_resume,
>  	.check_power_status = snd_hda_gen_check_power_status,
>  	.stream_pm = snd_hda_gen_stream_pm,
>  };
> -- 
> 2.25.1
>