Re: linux-next: Tree for June 20: [regression] No sound on SB600

[Takashi Iwai <tiwai@suse.de>]

At Sat, 21 Jun 2008 21:25:41 +0200,
Rafael J. Wysocki wrote:
> 
> On Saturday, 21 of June 2008, Takashi Iwai wrote:
> > At Sat, 21 Jun 2008 17:28:09 +0200,
> > Rafael J. Wysocki wrote:
> > > 
> > > On Friday, 20 of June 2008, Stephen Rothwell wrote:
> > > > Hi all,
> > > > 
> > > > Changes since next-20080619:
> > > > 
> > > > The net tree lost its conflict against Linus' tree.
> > > > 
> > > > The bkl-removal tree lost a conflict.
> > > 
> > > On one of my test boxes with the ATI chipset (SB600 south bridge) there's no
> > > sound output.  Bisection shows that the following commit causes this to happen
> > > and reverting the commit fixes the problem for me:
> > > 
> > > commit cf97e6678f1e1e7bbdb566c36771540123832a31
> > > Author: Takashi Iwai <tiwai@suse.de>
> > > Date:   Mon Jun 16 15:47:26 2008 +0200
> > > 
> > >     ALSA: hda - Fix wrong volumes in AD1988 auto-probe mode
> > >     
> > >     Don't create mixer volume elements for Headphone and Speaker if they
> > >     use the same DAC as normal line-outs on AD1988.  Otherwise the amp
> > >     value gets screwed up, e.g.
> > >     	https://bugzilla.novell.com/show_bug.cgi?id=398255
> > >     
> > >     Signed-off-by: Takashi Iwai <tiwai@suse.de>
> > > 
> > > Please let me know if you need any additional information.
> > 
> > Could you run alsa-info.sh with --no-upload option and attach the
> > output for both states with and without that commit?  The script is
> > found at
> > 	http://www.alsa-project.org/alsa-info.sh
> > Also, show the relevant kernel message from snd-hda-intel (informing
> > the auto-parsing of pins, etc), too.
> 
> dmesg outputs and the outputs of 'alsa-info.sh --no-upload' for the kernel with
> the patch applied (1) and with the reverted (2) are at:
> 
> http://www.sisk.pl/kernel/debug/20080620/box2/dmesg-1.log
> http://www.sisk.pl/kernel/debug/20080620/box2/alsa-info-1.txt
> 
> http://www.sisk.pl/kernel/debug/20080620/box2/dmesg-2.log
> http://www.sisk.pl/kernel/debug/20080620/box2/alsa-info-2.txt

Well, this is actually no regression but an unlucky side-effect of the
fix.

In the old version, "Front" volume didn't work properly because of
confliction with a bogus "Headphone" volume.  Now, a wrong volume
element is removed but only "Front" remains while "Front" wasn't set
up properly.

In short, the below should fix the problem:

	% amixer set Front 90%


thanks,

Takashi