Re: Problems with Realtek HD Audio

[Takashi Iwai <tiwai@suse.de>]

At Fri, 9 Feb 2007 15:25:12 +0900 (JST),
Shingo HORISAWA wrote:
> 
> Hi, Takashi.
> 
> 
> > Well, the codec is usually recognized, but it
> > doesn't mean that your
> > hardware is really supported.  Check out the kernel
> > log and see
> > whether the driver tells about "unknown model...". 
> > Also try to build
> > drivers with --with-debug=detect option for making
> > it more verbose.
> 
> I tried this option, but I couldn't find such messages (or
> any messages) in /var/log/messages.

Hm, but if built properly with --with-debug=detect option, the driver
must tells you a model name when you pass model=XXX option.  Make sure
that the model option is really passed.  You can see the actual value
in /sysfs/modules/snd_hda_intel/parameters/model file.


> > >    kernel: hda-intel: Invalid position buffer,
> > using LPIB
> > > read method instead.
> > 
> > Does it appear after arecord or even after aplay?
> 
> No, it only appeared at the start of arecord.
> 
> > I recommend you to try various model options for
> > ALC262.  There are
> > plenty of varients, but not specific for Toshiba, so
> > far.  But it's
> > still worth to try.
> 
> I also tried all model options in ALSA-configuration.txt,
> but it was the same result.
> 
> 
> But, I noticed three facts:
> 
>   a) This problem occurs when applications are started. 
>      But, after restart ALSA, it doesn't occur 
>      when I start the "same" application again.
>      (When I start another application, it occurs...)
> 
>   b) When I run "arecord" after some applications are
> started, 
>      this problem doesn't occur.
>      (When I started "new" application while I'm running
> "arecord", it occurs.)
> 
>   c) On CD-boot linux, such as KNOPPIX, 
>      this problem doesn't occur at all.
> 
> 
> I doubted that this is a problem around memory.
> So, to make sure that, I have tried to clean a page cache
> after this problem occured as follows:
> 
>   (1) run "arecord"
>   (2) start wireshark(ethereal)  => arecord stopped
>   (3) restart ALSA
>   (4) run "arecord" again => OK
>   (5) start wireshark => arecord is still running
>   (6) stop wireshark
>   (7) clean a page cache as follows:
>          # systcl -w vm.drop_caches=3 
>         => arecord is still running
>   (8) start wireshark again => arecord stopped!

I still have no idea why wireshark interferes the sound stuff.  Maybe
both are on the same IRQ line?  You can try enable_msi=1 option to
snd-hda-intel to split the HD-Audio over MSI, just for testing.

But, anyway, there are too little information from the sound driver
itself.  Usually when something wrong occurs, the driver spits
annoying warnings.  But in your case, it doesn't produce any errors.
Puzzling...


Takashi

-------------------------------------------------------------------------
Using Tomcat but need to do more? Need to support web services, security?
Get stuff done quickly with pre-integrated technology to make your job easier.
Download IBM WebSphere Application Server v.1.0.1 based on Apache Geronimo
http://sel.as-us.falkag.net/sel?cmd=lnk&kid=120709&bid=263057&dat=121642