No sound with Via 823x hardware on DFI K8M800-MLVF Motherboard with kernels 2.6.10-3/4

[Joe MacDonald]

I have a 2.6.10-3, 2.6.10-4 and a 2.6.8.1-5 kernel installed on my Hoary system
right now. If I choose to boot into one of the 2.6.10 kernels I have no sound
of any kind, though anything trying to access the sound hardware seems to think
everything is functioning properly (for example the Multimedia Systems Selector
'Default Sink' test button doesn't generate errors, but neither does it create
any sounds). If I choose to boot into the 2.6.8.1-5 kernel, however, everything
works correctly, sound functions normally.

Script logs of cat-ing the files in /proc/asound can be found at:

http://www.deserted.net/~joe/sound-find-broken.typescript
http://www.deserted.net/~joe/sound-find-working.typescript

[Joe MacDonald (joeythesaint)]

Created an attachment (id=1546)
Results of cat on all files in /proc/asound on a 2.6.10-3 kernel

[Joe MacDonald (joeythesaint)]

Created an attachment (id=1547)
Results of cat on all files in /proc/asound on a 2.6.8.1-5 kernel

[kaptk2 (andrew-pure-wireless)]

The work around I found for this bug involved setting capture monitor to mute
(in alsamixer) and muteing everything that was had surround in its name. This
looks like a bug in the inital ALSA setup. Pehaps change the defualt to have the
capture monitor muted.

[Joe MacDonald (joeythesaint)]

Weird. I don't know why playing with the input settings would have anything to
do with it, but when I muted those and the surround sound options I got my sound
back with the latest 2.6.10-5 build. Since I opened this one, I guess I'll
close it now that the problem is solved for me. Thanks for the help!

[Matt Zimmerman (mdz)]

If these settings are required in order to have the device work properly, then
we should arrange for them to be set by default. Please attach amixer output
and tell us the minimum settings which needed to be changed in order to get
things working.

[Joe MacDonald (joeythesaint)]

No problem. I just conducted an experiement where I turned on everything,
verified my sound wasn't working then turned off one thing at a time until it
started working again. It looks like the only thing I needed to mute was the
capture monitor (IEC958 Capture Monitor [off]). I'm attaching the output of
amixer, too. The entry in question is at line 85 of the file.

[Joe MacDonald (joeythesaint)]

Created an attachment (id=2072)
Results of amixer command showing functioning sound configuration for 2.6.10-5
kernel.

[Thomas Hood (jdthood)]

(In reply to comment #6)
> No problem. I just conducted an experiement where I turned on everything,
> verified my sound wasn't working then turned off one thing at a time until it
> started working again. It looks like the only thing I needed to mute was the
> capture monitor (IEC958 Capture Monitor [off]). I'm attaching the output of
> amixer, too. The entry in question is at line 85 of the file.

OK, I will patch the Debian alsa initscript with this:

--- /etc/init.d/alsa 2005-03-30 10:59:45.000000000 +0200
+++ alsa-base.alsa.init 2005-04-10 21:57:35.000000000 +0200
@@ -211,6 +211,9 @@
        switch_control "Headphone Jack Sense" off
        switch_control "Line Jack Sense" off

+ # Required at least for Via 823x hardware on DFI K8M800-MLVF Motherboard
with kernels 2.6.10-3/4 (see ubuntu #7286)
+ switch_control "IEC958 Capture Monitor" off
+
        return 0
 }

[Clint Chapman (clint-ubiquityhosting)]

*** Bug 15016 has been marked as a duplicate of this bug. ***

[Matt Zimmerman (mdz)]

In bug #15016, Daniel T. Chen stated:

As a side note, this was
fixed in upstream 1.0.9rc2 (see cmi9761 ac97 codec-related fixes at
http://cvs.sourceforge.net/viewcvs.py/alsa/alsa-kernel/pci/ac97/ac97_codec.c?rev=1.172&view=log).

so perhaps we don't need to work around this in alsa-base at all?

[Daniel T Chen (crimsun)]

(In reply to comment #10)
> so perhaps we don't need to work around this in alsa-base at all?

For 1.0.9, we will not have to work around this issue in alsa-base.

[Thomas Hood (jdthood)]

This bug has been fixed in the 1.0.9 alsa-base package, so I think that this bug report can now
be closed.

[Chuck Short (zulcss)]

Closing

[Daniel T Chen (crimsun)]

*** Bug 16699 has been marked as a duplicate of this bug. ***