Bug 229224 - snd-hda-intel does not work consistently
Summary: snd-hda-intel does not work consistently
Status: RESOLVED WONTFIX
: 192884 (view as bug list)
Alias: None
Product: openSUSE 10.2
Classification: openSUSE
Component: Sound (show other bugs)
Version: Final
Hardware: i386 Other
: P5 - None : Normal (vote)
Target Milestone: ---
Assignee: Takashi Iwai
QA Contact: E-mail List
URL:
Whiteboard:
Keywords:
Depends on:
Blocks:
 
Reported: 2006-12-16 21:01 UTC by Forgotten User l5QCzUHucG
Modified: 2008-06-25 09:23 UTC (History)
2 users (show)

See Also:
Found By: Other
Services Priority:
Business Priority:
Blocker: ---
Marketing QA Status: ---
IT Deployment: ---


Attachments
lspci (8.10 KB, text/plain)
2006-12-19 21:07 UTC, Forgotten User l5QCzUHucG
Details
dmesg (19.71 KB, text/plain)
2006-12-19 21:08 UTC, Forgotten User l5QCzUHucG
Details

Note You need to log in before you can comment on or make changes to this bug.
Description Forgotten User l5QCzUHucG 2006-12-16 21:01:09 UTC
I have a laptop with the following audio device:

00:1b.0 Audio device: Intel Corporation 82801G (ICH7 Family) High Definition Audio Controller (rev 02)
        Subsystem: Gateway 2000 Unknown device 0366
        Flags: bus master, fast devsel, latency 0, IRQ 21
        Memory at d8240000 (64-bit, non-prefetchable) [size=16K]
        Capabilities: [50] Power Management version 2
        Capabilities: [60] Message Signalled Interrupts: Mask- 64bit+ Queue=0/0 Enable-
        Capabilities: [70] Express Unknown type IRQ 0
        Capabilities: [100] Virtual Channel
        Capabilities: [130] Unknown (5)

which gets set up correctly using Yast2 and sound works fine.

Now after a number of reboots, it stops working, ie no start-up sound in KDE and no sound gets played when playing wav sounds through Xine. I then found that the Yast sound configuration module does not reliable re-enable it. I usually have to use alsaconf to re-enable it which works most times although not always, either.

Happy to provide any logs needed...
Comment 1 Marcus Schaefer 2006-12-18 11:06:51 UTC
sorry sound is not my playground
Comment 3 Takashi Iwai 2006-12-19 14:13:08 UTC
Please check the kernel log after suspend.

I guess simply reloading the driver via 'rcalsasound restart' would repair the sound, too.

Is it suspend to RAM or to DISK?
Comment 4 Takashi Iwai 2006-12-19 14:13:49 UTC
Also I need every detail of your laptop, as much as possible.
Comment 5 Forgotten User l5QCzUHucG 2006-12-19 21:06:11 UTC
Takashi,

This happens after a normal startup, no suspend to disk or RAM involved. As it turns out, the Xen kernel seems to be much more consistent with the alsa driver then the non-Xen kernel. I also added my user name to the audio group after having read various threads about the potential benefits of it on the opensuse mailing list. However, just now I booted into the standard kernel and sound was not working. It started working with rcalsasound restart but after calling that a few times, it stopped again and no matter what I tried, I couldn't get it to work again. I then rebooted into the xen-kernel and voila, I have sound again.

I'll attach the output of lspci -v and dmesg which I hope is as much information as you need. In short it's a Gateway lapto with Intel Duo T2500 processor,945GM and 2GB of RAM. 
Comment 6 Forgotten User l5QCzUHucG 2006-12-19 21:07:38 UTC
Created attachment 110419 [details]
lspci
Comment 7 Forgotten User l5QCzUHucG 2006-12-19 21:08:29 UTC
Created attachment 110420 [details]
dmesg
Comment 8 Forgotten User l5QCzUHucG 2006-12-20 14:46:52 UTC
I probably jinxed it now with the Xen kernel as it just stopped working with that kernel, too. And this time the sound remains silent even after rcalsasound restart and alsaconf...
Comment 9 Takashi Iwai 2006-12-20 15:13:34 UTC
I don't see any critical error messages in the attached kernel log, so it's more likely an application side error.  (Most of HD-audio errors come from IRQ mishandling, resulting in 'azx_get_response timeout' messages.)

Make sure that the mixer is unmuted and adjust when it happens.
Try a simple program as tests, such as
    % aplay -vv /usr/share/sounds/alsa/test.wav
Comment 10 Forgotten User l5QCzUHucG 2006-12-20 20:35:13 UTC
Takashi,

The mixer settings are actually quite interesting in this respect. Whenever the sound works Kmix shows only two channels (Master and PCM with both having a green light) and when it does not work Kmix shows 3 channels (Master and PCM and SPDIF). In the latter case Master and SPDIF have the green LED whereas PCM does not. I reckon that's the cause. As a case in point, I just restarted my laptop and had sound. I then called rcalsasound restart and lost the sound. I tried running alsaconf but that didn't restore the sound either. /var/log/messages was not particularly illuminating either:
<rcalsasound restart>

Dec 20 20:22:56 mobilix kernel: PCI: Enabling device 0000:00:1b.0 (0100 -> 0102)
Dec 20 20:22:56 mobilix kernel: ACPI: PCI Interrupt 0000:00:1b.0[A] -> GSI 22 (level, low) -> IRQ 225
Dec 20 20:22:56 mobilix kernel: PCI: Setting latency timer of device 0000:00:1b.0 to 64
Dec 20 20:28:55 mobilix su: (to root) swesemeyer on /dev/pts/2

<alsaconf>
Dec 20 20:29:01 mobilix kernel: ACPI: PCI interrupt for device 0000:00:1b.0 disabled
Dec 20 20:29:27 mobilix kernel: PCI: Enabling device 0000:00:1b.0 (0100 -> 0102)
Dec 20 20:29:27 mobilix kernel: ACPI: PCI Interrupt 0000:00:1b.0[A] -> GSI 22 (level, low) -> IRQ 225
Dec 20 20:29:27 mobilix kernel: PCI: Setting latency timer of device 0000:00:1b.0 to 64

Does that shed any more light on the situation? Do you need any output from any of the files in /proc/asound?
Comment 11 Forgotten User l5QCzUHucG 2006-12-21 16:09:28 UTC
Just to clarify the above comment. There is no "LED" at all above the PCM channel, so I cannot mute/unmute it at all and moving the slider does not have any effect at all.
Comment 12 Forgotten User l5QCzUHucG 2007-01-02 21:15:43 UTC
Not sure whether this adds anything but here is the output of
cat /proc/asound/card0/codec#0|head
Codec: SigmaTel ID 7634
Address: 0
Vendor Id: 0x83847634
Subsystem Id: 0x107b0461
Revision Id: 0x100101
Default PCM: rates 0x7e0, bits 0x0e, types 0x1
Default Amp-In caps: N/A
Default Amp-Out caps: ofs=0x1f, nsteps=0x1f, stepsize=0x05, mute=1
Node 0x02 [Audio Output] wcaps 0xd0401: Stereo
  Power: 0x0

It looks as if this bug is not only present in 10.2 but also in 10.1:
http://www.linuxquestions.org/questions/showthread.php?t=473423

It has also been reported on:
http://www.mail-archive.com/alsa-user@lists.sourceforge.net/msg17745.html

Comment 13 Forgotten User l5QCzUHucG 2007-01-08 21:56:50 UTC
in addition to comment 12, there is a difference between the contents of codec#0 when the sound is working and when it is not working
the bit that differs is in the following section:

working:
Node 0x06 [Audio Selector] wcaps 0x300901: Stereo
 Connection: 3
    0x02* 0x07 0x14

not-working:
Node 0x06 [Audio Selector] wcaps 0x300901: Stereo
 Connection: 3
    0x02 0x07* 0x14

(also see http://www.mail-archive.com/alsa-user@lists.sourceforge.net/msg17777.html)

So is there any way to force it to use the settings that work?
Comment 14 Forgotten User l5QCzUHucG 2007-01-09 20:31:57 UTC
Takashi,

Does this bug report

https://bugtrack.alsa-project.org/alsa-bug/view.php?id=2146

apply to my bug, too?

It looks pretty much like what I found. I guess that'll mean it's going to be fixed in 10.3?
Comment 15 Takashi Iwai 2007-01-16 11:29:21 UTC
Well, 2.6.20 kernel contains only a part of ALSA 1.0.14rc*.  So, it's possible that it's still unfixed.  But anyway, you can try the latest kernel package on FACTORY, which is already based on the latest 2.6.20-rc.
Comment 16 Forgotten User l5QCzUHucG 2007-01-27 16:24:53 UTC
I downloaded 1.0.14rc2 yesterday, compiled and installed it and now I have shedloads of channels and 3 tabs in KMIX and what's more, sounds seems to be working fine so far.
I guess once 1.0.14 is part of the official OpenSUSE kernels this bug can be closed.
Comment 17 Will Stephenson 2007-01-30 08:15:11 UTC
*** Bug 192884 has been marked as a duplicate of this bug. ***
Comment 18 Takashi Iwai 2007-02-02 14:49:29 UTC
Unfortunately, it'll take some time that all portion of ALSA 1.0.14 is taken into the upstream.  So far, only a part of them is in 2.6.20 kernel currently used for FACTORY.
Comment 19 Takashi Iwai 2007-02-26 12:05:56 UTC
The fix is on upstream, so resolved to LATER for now.
Comment 20 Stephan Kulow 2008-06-25 09:19:36 UTC
mass reopening all 10.2 LATER+REMIND bugs.
Comment 21 Stephan Kulow 2008-06-25 09:23:53 UTC
close all 10.2 LATER/REMIND bugs as WONTFIX. Reopen yourself if you still plan to work on it.