Bug#658728: linux-image-3.2.0-1-amd64: No more sound
A. Costa wrote:
> Thanks, and sorry for the mis-filing. Unfortunately I'm not up on
> kernel packaging, so could you therefore advise where to file this?
The usual way would be "reportbug linux-image-$(uname -r)". No need
for that this time --- I've cloned the bug as bug#658728, so please
direct followups there.
[...]
> So perhaps this bug, if it's only an Asus laptop thing, might be one
> instance of a greater problem.
Sorry for the lack of clarity. These symptoms are indeed expected to
affect many laptops with an ALC861 codec, not just Asus laptops. The
problem is that different machines use a different mechanism to turn
on or off sound completely. The driver has a table that documents
this sort of thing for (1) all ASUS laptops (after the fix to
bug#657302), (2) Haier W18, (3) FSC Amilo Pi1505.
So, the next step is to work with upstream to get the pin routing
information for your laptop. I'll send some hints on this in a
separate message.
Thanks,
Jonathan
--
To UNSUBSCRIBE, email to debian-kernel-REQUEST@lists.debian.org
with a subject of "unsubscribe". Trouble? Contact listmaster@lists.debian.org
Archive: 20120205152156.GD32584@burratino">http://lists.debian.org/20120205152156.GD32584@burratino
02-06-2012, 01:42 AM
"A. Costa"
Bug#658728: linux-image-3.2.0-1-amd64: No more sound
On Sun, 5 Feb 2012 09:21:56 -0600
Jonathan Nieder <jrnieder@gmail.com> wrote:
> Sorry for the lack of clarity. These symptoms are indeed expected to
> affect many laptops with an ALC861 codec, not just Asus laptops. The
> problem is that different machines use a different mechanism to turn
> on or off sound completely. The driver has a table that documents
> this sort of thing for (1) all ASUS laptops (after the fix to
> bug#657302), (2) Haier W18, (3) FSC Amilo Pi1505.
>
> So, the next step is to work with upstream to get the pin routing
> information for your laptop. I'll send some hints on this in a
> separate message.
It might not matter, but my current system, correctly identified by
'alsa-info.sh' as an "HP Compaq dx2200 MT", is a Desktop box, not a
laptop. The sound card is part of the mobo, however, which is somewhat
laptop-like. More here:
Manuals (guides, supplements, addendums, etc)
HP Compaq dx2200 Microtower PC
http://h20000.www2.hp.com/bizsupport/TechSupport/DocumentIndex.jsp?contentType=SupportManual&lang=e n&cc=us&prodSeriesId=1844701&prodTypeId=12454
To fetch pin routing data via 'hda-verb' requires recompiling the
kernel. This week I haven't enough time to build a kernel.
(Bandwidth=dialup.) Pity there isn't a simpler method; even using a
multitester on the mobo would be less hassle than rebuilding the kernel.
Searching the web for 'HP Compaq dx2200 MT "hda-verb"' finds some kernel module patch notes from 2007:
[ALSA] hda-codec - Add quirks for HP dx2200/dx2250
HP dx2200 and dx2250 use Micro-Star International (MSI) motherboards
(models MS-7254 and MS-7297 respectively) with an ALC862 codec in
threestack configuration. Adding this quirk allows correct 5.1 sound
output in these systems.
Sound on this box with my previous stock Debian kernels worked. Today
I reverted to 'linux-image-3.1.0-1-686-pae', under which the
sound still works fine. Attached are the outputs of 'alsa-info.sh' for
the the downgraded (good sound 3.1.0-1) kernel, and the newer (no sound
working 3.2.0-1) kernel.
As advised I've CC:'d this email upstream, for whom the relevant Debian
bug reports can be found here:
Bug#658728: linux-image-3.2.0-1-amd64: No more sound
At Sun, 5 Feb 2012 21:42:28 -0500,
A. Costa wrote:
>
> On Sun, 5 Feb 2012 09:21:56 -0600
> Jonathan Nieder <jrnieder@gmail.com> wrote:
>
> > Sorry for the lack of clarity. These symptoms are indeed expected to
> > affect many laptops with an ALC861 codec, not just Asus laptops. The
> > problem is that different machines use a different mechanism to turn
> > on or off sound completely. The driver has a table that documents
> > this sort of thing for (1) all ASUS laptops (after the fix to
> > bug#657302), (2) Haier W18, (3) FSC Amilo Pi1505.
> >
> > So, the next step is to work with upstream to get the pin routing
> > information for your laptop. I'll send some hints on this in a
> > separate message.
>
> It might not matter, but my current system, correctly identified by
> 'alsa-info.sh' as an "HP Compaq dx2200 MT", is a Desktop box, not a
> laptop. The sound card is part of the mobo, however, which is somewhat
> laptop-like. More here:
>
> Manuals (guides, supplements, addendums, etc)
> HP Compaq dx2200 Microtower PC
> http://h20000.www2.hp.com/bizsupport/TechSupport/DocumentIndex.jsp?contentType=SupportManual&lang=e n&cc=us&prodSeriesId=1844701&prodTypeId=12454
>
> To fetch pin routing data via 'hda-verb' requires recompiling the
> kernel. This week I haven't enough time to build a kernel.
> (Bandwidth=dialup.) Pity there isn't a simpler method; even using a
> multitester on the mobo would be less hassle than rebuilding the kernel.
>
> Searching the web for 'HP Compaq dx2200 MT "hda-verb"' finds some kernel module patch notes from 2007:
>
> [ALSA] hda-codec - Add quirks for HP dx2200/dx2250
>
> HP dx2200 and dx2250 use Micro-Star International (MSI) motherboards
> (models MS-7254 and MS-7297 respectively) with an ALC862 codec in
> threestack configuration. Adding this quirk allows correct 5.1 sound
> output in these systems.
>
> Signed-off-by: Claudio Matsuoka <cmatsuoka@gmail.com>
> Signed-off-by: Takashi Iwai <tiwai@suse.de>
> Signed-off-by: Jaroslav Kysela <perex@suse.cz>
> http://www.codebrowse.net/history/linux-3.0/sound/pci/hda/patch_realtek.c
>
> Sound on this box with my previous stock Debian kernels worked. Today
> I reverted to 'linux-image-3.1.0-1-686-pae', under which the
> sound still works fine. Attached are the outputs of 'alsa-info.sh' for
> the the downgraded (good sound 3.1.0-1) kernel, and the newer (no sound
> working 3.2.0-1) kernel.
>
> As advised I've CC:'d this email upstream, for whom the relevant Debian
> bug reports can be found here:
>
> http://bugs.debian.org/cgi-bin/bugreport.cgi?bug=657302&archived=False&mbox=no
> http://bugs.debian.org/cgi-bin/bugreport.cgi?bug=658728&archived=False&mbox=no
No sound from which output exactly?
What happens if you turn off "Auto-Mute Mode" mixer enum?
% amixer -c0 set "Auto-Mute Mode" Disabled
This is the control to turn on/off the automatic mute of the speakers
via line-out or headphone jack plugs. It's possible that this doesn't
work expectedly when the hardware implementation doesn't follow the
standard.
Looking through both alsa-info.sh outputs, your case doesn't seem like
a NID-0x0f VREF issue like ASUS laptops. Even 3.1.x kernel shows 0xc0
(i.e. without VREF) for the pin-control of NID 0x0f. The problem is
likely different.
thanks,
Takashi
--
To UNSUBSCRIBE, email to debian-kernel-REQUEST@lists.debian.org
with a subject of "unsubscribe". Trouble? Contact listmaster@lists.debian.org
Archive: s5hsjioqpuj.wl%tiwai@suse.de">http://lists.debian.org/s5hsjioqpuj.wl%tiwai@suse.de
02-07-2012, 02:53 AM
"A. Costa"
Bug#658728: linux-image-3.2.0-1-amd64: No more sound
On Mon, 06 Feb 2012 10:21:40 +0100
Takashi Iwai <tiwai@suse.de> wrote:
> No sound from which output exactly?
Which? None. No outputs are audible, so far as I can tell. The
'pavucontrol' streams still work, the meter goes up and down to the
silent music.
On the good kernel (3.1.0-1) sound plays through the (separately
powered) speakers plugged into the back of the PC case; when I plug the
headphones in a different jack on the front of the case, the speakers
turn off, and the headphone sound goes on.
On the silent kernel (3.2.0-1), there's no sound from headphones or
speaker, and the headphones being plugged and unplugged has no audible
effect on either speakers or headphones.
> What happens if you turn off "Auto-Mute Mode" mixer enum?
> % amixer -c0 set "Auto-Mute Mode" Disabled
On the good kernel (3.1.0-1), this happens:
% amixer -c0 set "Auto-Mute Mode" Disabled ; echo $?
amixer: Unable to find simple control 'Auto-Mute Mode',0
1
Haven't tried that 'amixer' command on the silent kernel (3.2.0-1).
--
To UNSUBSCRIBE, email to debian-kernel-REQUEST@lists.debian.org
with a subject of "unsubscribe". Trouble? Contact listmaster@lists.debian.org
Archive: 20120206225359.0797f4c1804e1882aa3b3a3c@gis.net">h ttp://lists.debian.org/20120206225359.0797f4c1804e1882aa3b3a3c@gis.net
02-07-2012, 08:17 AM
Takashi Iwai
Bug#658728: linux-image-3.2.0-1-amd64: No more sound
At Mon, 6 Feb 2012 22:53:59 -0500,
A. Costa wrote:
>
> On Mon, 06 Feb 2012 10:21:40 +0100
> Takashi Iwai <tiwai@suse.de> wrote:
>
> > No sound from which output exactly?
>
> Which? None.
"None" means almost nothing Tell which outputs you have tried.
I don't know exactly whether you checked only the speaker, or only the
headphone, or any jacks what ever. (Judging from the text, I can
guess you tested the headphone jack and the line-out jack. But you
didn't test the surround/CLFE outputs from line-in/mic-in jacks, no?)
> No outputs are audible, so far as I can tell. The
> 'pavucontrol' streams still work, the meter goes up and down to the
> silent music.
>
> On the good kernel (3.1.0-1) sound plays through the (separately
> powered) speakers plugged into the back of the PC case; when I plug the
> headphones in a different jack on the front of the case, the speakers
> turn off, and the headphone sound goes on.
Hm, so maybe the auto-mute feature is implemented in the hardware
itself? Interesting. The 3.1 driver doesn't provide the auto-mute in
software for your device.
> On the silent kernel (3.2.0-1), there's no sound from headphones or
> speaker, and the headphones being plugged and unplugged has no audible
> effect on either speakers or headphones.
>
> > What happens if you turn off "Auto-Mute Mode" mixer enum?
> > % amixer -c0 set "Auto-Mute Mode" Disabled
>
> On the good kernel (3.1.0-1), this happens:
>
> % amixer -c0 set "Auto-Mute Mode" Disabled ; echo $?
> amixer: Unable to find simple control 'Auto-Mute Mode',0
>
> 1
>
> Haven't tried that 'amixer' command on the silent kernel (3.2.0-1).
The question above was only for 3.2 kernel. It wasn't enabled for 3.1
kernel for your device unless you passed model=auto explicitly.
If the above doesn't change anything, try to set the pin-control of
each pin as same as 3.1 kernel:
One of the command may trigger something. Don't change the plug while
you test this. Also, keep "Auto-Mute" mode turned off. Otherwise the
driver might reset the pin-control values again.
Last but not least, for testing the output, don't use PulseAudio but
use aplay or speaker-test with the raw ALSA access.
Run like
% aplay -Dplughw -vv somefile.wav
If PA complains, use pasuspender
% pasuspender -- aplay -Dplughw -vv somefile.wav
thanks,
Takashi
--
To UNSUBSCRIBE, email to debian-kernel-REQUEST@lists.debian.org
with a subject of "unsubscribe". Trouble? Contact listmaster@lists.debian.org
Archive: s5hy5sfovda.wl%tiwai@suse.de">http://lists.debian.org/s5hy5sfovda.wl%tiwai@suse.de
02-08-2012, 01:16 AM
"A. Costa"
Bug#658728: linux-image-3.2.0-1-amd64: No more sound
On Tue, 07 Feb 2012 10:17:37 +0100
Takashi Iwai <tiwai@suse.de> wrote:
> > On the good kernel (3.1.0-1), this happens:
> >
> > % amixer -c0 set "Auto-Mute Mode" Disabled ; echo $?
> > amixer: Unable to find simple control 'Auto-Mute Mode',0
> >
> > 1
> >
> > Haven't tried that 'amixer' command on the silent kernel (3.2.0-1).
>
> The question above was only for 3.2 kernel. It wasn't enabled for 3.1
> kernel for your device unless you passed model=auto explicitly.
Success:
% cat /proc/version
Linux version 3.2.0-1-686-pae (Debian 3.2.4-1) (waldi@debian.org) (gcc version 4.6.2 (Debian 4.6.2-12) ) #1 SMP Sun Feb 5 23:52:49 UTC 2012
Sound silently playing in the background (via 'audacious' and
'pulseaudio') suddenly went on. The headphones when plugged in also
worked correctly.
Exactly which jacks were tested before: the line out (the main stereo
only), and the headphone jack. Did not test surround, line in, or the
mic. (I don't ordinarily use those jacks.)
At present the additional tests you suggested, ('hda-verb', 'aplay', and
'pasuspender'), might not be necessary. If those would still be useful,
please say so.
HTH, and thanks for coding & maintaining...
--
To UNSUBSCRIBE, email to debian-kernel-REQUEST@lists.debian.org
with a subject of "unsubscribe". Trouble? Contact listmaster@lists.debian.org
Archive: 20120207211657.d229eb19145eba785bac7b56@gis.net">h ttp://lists.debian.org/20120207211657.d229eb19145eba785bac7b56@gis.net
02-10-2012, 09:58 AM
Takashi Iwai
Bug#658728: linux-image-3.2.0-1-amd64: No more sound
At Tue, 7 Feb 2012 21:16:57 -0500,
A. Costa wrote:
>
> On Tue, 07 Feb 2012 10:17:37 +0100
> Takashi Iwai <tiwai@suse.de> wrote:
>
> > > On the good kernel (3.1.0-1), this happens:
> > >
> > > % amixer -c0 set "Auto-Mute Mode" Disabled ; echo $?
> > > amixer: Unable to find simple control 'Auto-Mute Mode',0
> > >
> > > 1
> > >
> > > Haven't tried that 'amixer' command on the silent kernel (3.2.0-1).
> >
> > The question above was only for 3.2 kernel. It wasn't enabled for 3.1
> > kernel for your device unless you passed model=auto explicitly.
>
> Success:
>
> % cat /proc/version
> Linux version 3.2.0-1-686-pae (Debian 3.2.4-1) (waldi@debian.org) (gcc version 4.6.2 (Debian 4.6.2-12) ) #1 SMP Sun Feb 5 23:52:49 UTC 2012
>
> % amixer -c0 set "Auto-Mute Mode" Disabled ; echo $?
> Simple mixer control 'Auto-Mute Mode',0
> Capabilities: enum
> Items: 'Disabled' 'Enabled'
> Item0: 'Disabled'
> 0
>
> Sound silently playing in the background (via 'audacious' and
> 'pulseaudio') suddenly went on. The headphones when plugged in also
> worked correctly.
Do you mean that the speaker is still muted automatically with the
headphone plugging, even though "Auto-Mute Mode" mixer element is
disabled? Or the speaker keeps playing?
> Exactly which jacks were tested before: the line out (the main stereo
> only), and the headphone jack. Did not test surround, line in, or the
> mic. (I don't ordinarily use those jacks.)
>
> At present the additional tests you suggested, ('hda-verb', 'aplay', and
> 'pasuspender'), might not be necessary. If those would still be useful,
> please say so.
If disabling auto-mute mode fixes the thing, I'll create a blacklist
and add this device to it so that this control won't be created in
future.
thanks,
Takashi
--
To UNSUBSCRIBE, email to debian-kernel-REQUEST@lists.debian.org
with a subject of "unsubscribe". Trouble? Contact listmaster@lists.debian.org
Archive: s5h1uq35509.wl%tiwai@suse.de">http://lists.debian.org/s5h1uq35509.wl%tiwai@suse.de
02-10-2012, 10:57 PM
"A. Costa"
Bug#658728: linux-image-3.2.0-1-amd64: No more sound
On Fri, 10 Feb 2012 11:58:46 +0100
Takashi Iwai <tiwai@suse.de> wrote:
> > Sound silently playing in the background (via 'audacious' and
> > 'pulseaudio') suddenly went on. The headphones when plugged in also
> > worked correctly.
>
> Do you mean that the speaker is still muted automatically with the
> headphone plugging, even though "Auto-Mute Mode" mixer element is
> disabled? Or the speaker keeps playing?
When plugged in, the headphones played and the speaker muted.
Unplugging the phones, the speakers play.
At the risk of being redundant, here's what I have so far:
First off, with plain kernel v3.2.0-1-686-pae, I run 'audacious' with a
playlist. 'pavucontrol' shows movement. The speakers are mute.
Plugged in, the headphone are also mute.
Then I ran 'amixer -c0 set "Auto-Mute Mode" Disabled'.
The speakers (line out) play. Plug in the headphones, the speaker
mutes, the headphones play. Unplug the phones, the speakers play.
HTH...
--
To UNSUBSCRIBE, email to debian-kernel-REQUEST@lists.debian.org
with a subject of "unsubscribe". Trouble? Contact listmaster@lists.debian.org
Archive: 20120210185734.c021b12f7bae08aa5e311701@gis.net">h ttp://lists.debian.org/20120210185734.c021b12f7bae08aa5e311701@gis.net
02-13-2012, 10:10 AM
Takashi Iwai
Bug#658728: linux-image-3.2.0-1-amd64: No more sound
At Fri, 10 Feb 2012 18:57:34 -0500,
A. Costa wrote:
>
> On Fri, 10 Feb 2012 11:58:46 +0100
> Takashi Iwai <tiwai@suse.de> wrote:
>
> > > Sound silently playing in the background (via 'audacious' and
> > > 'pulseaudio') suddenly went on. The headphones when plugged in also
> > > worked correctly.
> >
> > Do you mean that the speaker is still muted automatically with the
> > headphone plugging, even though "Auto-Mute Mode" mixer element is
> > disabled? Or the speaker keeps playing?
>
> When plugged in, the headphones played and the speaker muted.
> Unplugging the phones, the speakers play.
>
> At the risk of being redundant, here's what I have so far:
>
> First off, with plain kernel v3.2.0-1-686-pae, I run 'audacious' with a
> playlist. 'pavucontrol' shows movement. The speakers are mute.
> Plugged in, the headphone are also mute.
>
> Then I ran 'amixer -c0 set "Auto-Mute Mode" Disabled'.
> The speakers (line out) play. Plug in the headphones, the speaker
> mutes, the headphones play. Unplug the phones, the speakers play.
Thanks for confirmation.
I fixed the upstream code no to create the superfluous Auto-Mute
control. But the change is a bit intrusive so it's targeted only for
3.4 kernel, as you can work around it easily by turning off Auto-Mute
even in the current tree.
Takashi
--
To UNSUBSCRIBE, email to debian-kernel-REQUEST@lists.debian.org
with a subject of "unsubscribe". Trouble? Contact listmaster@lists.debian.org
Archive: s5hfwef2dlm.wl%tiwai@suse.de">http://lists.debian.org/s5hfwef2dlm.wl%tiwai@suse.de
02-15-2012, 07:57 PM
Jonathan Nieder
Bug#658728: linux-image-3.2.0-1-amd64: No more sound
tags 658728 + patch moreinfo
quit
Takashi Iwai wrote:
> I fixed the upstream code no to create the superfluous Auto-Mute
> control. But the change is a bit intrusive so it's targeted only for
> 3.4 kernel, as you can work around it easily by turning off Auto-Mute
> even in the current tree.
Thanks.
A. Costa, could you try the attached patches, against 3.2.y? It works
like this:
Add a new flag to indicate that the codec has no jack-detection cap.
This flag should be set for hardwares that have no jack-detect
implementation although the codec chip itself supports it.
A few machines with ALC861 & co are reported not to work properly with
the auto-mute feature in software. The auto-mute feature is implemented
in the hardware level, and the jack-detection never works with them.
Also, rename the fixup index as ALC861_FIXUP_* to follow the standard.
[jn: the original rationalized the name of the fixup index, but for
3.2.y let's not]
/* On some laptops, VREF of pin 0x0f is abused for controlling the main amp */
@@ -5256,6 +5258,14 @@ static void alc861_fixup_asus_amp_vref_0f(struct hda_codec *codec,
spec->keep_vref_in_automute = 1;
}
Bug#658728: linux-image-3.2.0-1-amd64: No more sound
Tell which outputs you have tried.
