Bug 203831

Created attachment 283147 [details]
dmesg of resuming from sleep and having the issue.

After some long spend hours its either 
ALSA: hda/realtek - Improve the headset mic for Acer Aspire laptops
or ALSA: hda/realtek - Check headset type by unplug and resume.
And the name of the bug seems to be incorrect since this seems to only occur on Acer Nitro 5 or maybe other Acer laptops.
The commit on mainline is c5ba1712661233ce0f4666b8c3dee5bb78d380f2.
If somebody could look at this so that we could get a fix in before full 5.2 release ?

The most suspected one is the commit aeac1a0dadb4ab699a61136a5729d048abe6ee5f
    ALSA: hda/realtek - Check headset type by unplug and resume

Could you just revert it and see whether the problem is gone?

(In reply to Takashi Iwai from comment #3)
> The most suspected one is the commit aeac1a0dadb4ab699a61136a5729d048abe6ee5f
>     ALSA: hda/realtek - Check headset type by unplug and resume
> 
> Could you just revert it and see whether the problem is gone?

Will report once the build is done.

(In reply to Takashi Iwai from comment #3)
> The most suspected one is the commit aeac1a0dadb4ab699a61136a5729d048abe6ee5f
>     ALSA: hda/realtek - Check headset type by unplug and resume
> 
> Could you just revert it and see whether the problem is gone?

Reverting that commit doesnt fix the issue.

Reverting commit ALSA: hda/realtek - Improve the headset mic for Acer Aspire laptops fixes the issue.

OK, then could you give alsa-info.sh output?  Run it with --no-upload option and attach the output to Bugzilla.

(In reply to Takashi Iwai from comment #7)
> OK, then could you give alsa-info.sh output?  Run it with --no-upload option
> and attach the output to Bugzilla.

Of the broken kernel or the fixed one ?
I guess broken one so i will have to build it

Created attachment 283257 [details]
alsa-info not working sound

Created attachment 283259 [details]
alsa-info working sound

Here are both for not working sound (quiet and when high scratching) and replugged the headphones and rerun it with working sound.

Thanks.

Hui, your patch seems causing a regression on another Acer model.
Could you check it?  In the worst case, we need to revert it before the 5.2 final release.

Hello Haxk20,

The patch (ALSA: hda/realtek - Improve the headset mic for Acer Aspire laptops) just changes the headset-mic from phantom audio jack to normal audio jack. Without this patch, the external mic is a phantom jack, it is always true, then it always shows up in the gnome UI, users can select it anytime.

After applying this patch, the external mic is a normal jack, if users plug a headset (headphone+mic) into the jack, the mic will shows up in the gnome UI; if users plug a headphone into the jack, the mic will not show up in the gnome UI.

So after applying this patch and plugging a headset into the jack, does the mic Jack change to true like below (in the alsa-info.txt):

	control.17 {
		iface CARD
		name 'Mic Jack'
		value true
		comment {
			access read
			type BOOLEAN
			count 1
		}
	}

If it doesn't change to true, we should revert this patch.

After plugging in earphones without microphone the control.17 remains false. I do not have earphones with microphone to test if that will change in any way.

I found some old earphones with microphone and it changed to true. So this behaviur is intended ?

OK but weird is that when i have the microphone earphones plugged in i can only see internal microphone in GNOME but it gets the audio from the earphones microphone. Shoudlnt there be option to select between external and internal microphone in gnome when user plugs them in ?

But still the bug remains that when this patch is applied and i restart or suspend PC with earphones plugged in i can hear audio very slightly but its scratching a little. This doesnt happen without this patch.

Reply the comment 16:
If the UI is (gnome-sound-setting + pulseaudio), when there is no external mic, the internal microphone will show up in the gnome-sound-setting, after plugging a earphone with mic, the internal microphone will disappear and the microphone will show up, then users only can record sound via earphone mic. This is the design by pulseaudio. So for your case, the Mic Jack changes to true, but there is still only internal microphone in the UI, I really don't have idea for it. Probably your UI is not (gnome-sound-setting + pulseaudio)?

Reply the comment 17:
For the noise issue, in theory I didn't touch spec->shutup/init_hook, it will not introduce noise for reboot or suspend/resume. I just wrote an email to kailang@realtek, let us see if he have some idea on it.

If we can't resolve the noise issue, let us revert this patch then.

Hello Haxk20,

Kailang and I had a short discussion about the noise, we all don't have any idea for this issue.

So please double confirm that after reverting the patch, there is no noise when rebooting and suspending, while after applying the patch, there is significant noise that most users can't ignore it. Then let us revert this patch.

Thanks

(In reply to Hui Wang from comment #19)
> Hello Haxk20,
> 
> Kailang and I had a short discussion about the noise, we all don't have any
> idea for this issue.
> 
> So please double confirm that after reverting the patch, there is no noise
> when rebooting and suspending, while after applying the patch, there is
> significant noise that most users can't ignore it. Then let us revert this
> patch.
> 
> Thanks

Hello,
Yes i can confirm for 100% that after reverting this patch there is no issue with sound.
After applying it tho the issue is that when you reboot or suspend with headphones plugged in and you play music after that or just any sound its not heard. Its only heard if the sound is of high volume but its scratching like crazy. This is impossible to ignore.
Thank you.

(In reply to Hui Wang from comment #18)
> Reply the comment 16:
> If the UI is (gnome-sound-setting + pulseaudio), when there is no external
> mic, the internal microphone will show up in the gnome-sound-setting, after
> plugging a earphone with mic, the internal microphone will disappear and the
> microphone will show up, then users only can record sound via earphone mic.
> This is the design by pulseaudio. So for your case, the Mic Jack changes to
> true, but there is still only internal microphone in the UI, I really don't
> have idea for it. Probably your UI is not (gnome-sound-setting + pulseaudio)?
> 
> Reply the comment 17:
> For the noise issue, in theory I didn't touch spec->shutup/init_hook, it
> will not introduce noise for reboot or suspend/resume. I just wrote an email
> to kailang@realtek, let us see if he have some idea on it.
> 
> If we can't resolve the noise issue, let us revert this patch then.

My UI is pulseaudio and gnome-sound-settings.

(In reply to Hui Wang from comment #18)
> Reply the comment 16:
> If the UI is (gnome-sound-setting + pulseaudio), when there is no external
> mic, the internal microphone will show up in the gnome-sound-setting, after
> plugging a earphone with mic, the internal microphone will disappear and the
> microphone will show up, then users only can record sound via earphone mic.
> This is the design by pulseaudio. So for your case, the Mic Jack changes to
> true, but there is still only internal microphone in the UI, I really don't
> have idea for it. Probably your UI is not (gnome-sound-setting + pulseaudio)?
> 
> Reply the comment 17:
> For the noise issue, in theory I didn't touch spec->shutup/init_hook, it
> will not introduce noise for reboot or suspend/resume. I just wrote an email
> to kailang@realtek, let us see if he have some idea on it.
> 
> If we can't resolve the noise issue, let us revert this patch then.

Im so sorry. Indeed the mic changes to external in the UI but this still doesnt fix the scratching sound.

The reverting patch is just sent out to review.

thx.

(In reply to Hui Wang from comment #23)
> The reverting patch is just sent out to review.
> 
> thx.

Thank you very much.
Since the patch is sent to review i will close this bug.