[alsa-devel] Bug report - patch_realtek.c - Laptop HP COMPAQ B1900 Series

[Raymond Yau]

2012/2/11, Jonathan Woithe <jwoithe at just42.net>:
> Hi Joey
>
> For future reference, please email reports like this to the alsa-devel list.
> The relevant developers are far more likely to respond to messages on the
> list than they are to messages sent privately.
>
> I have copied this to the list to maximise its distribution.
>
> On Sat, Feb 11, 2012 at 09:17:05AM +0800, joey.jiaojg wrote:
>> I'm reporting a bug related to ALSA and ALC260, which perhaps only
>> you can solve. Appreciate if you can check.
>> We got a lot of users using HP COMPAQ B1900 series which using ATI
>> SB450 and ALC260 for linux systems like Debian, Ubuntu and etc. The
>> problem is that there is no sound from speaker while enable
>> model=will can hear from earphone. The problem and alsa-info.txt is
>> reported here
>> https://answers.launchpad.net/ubuntu/+source/alsa-driver/+question/186698,
>> which thought be a bug of ALC260 driver bug.
>> Hope you can check what we can do to solve it.
>
> From the above I am not entirely sure what the problem is.  You seem to be
> talking about two separate conditions.  In the first (where I presume the
> module is loaded without any parameters you say there's no sound from the
> speaker.  Do other aspects of the card work correctly in this case?
>
> The second situation is where "model=will" is specified as a module
> parameter.  Here you say you can get audio from the headphone, but again
> what about other aspects of the soundcard - do they work normally?
>
> At a rough guess, it sounds like there are two underlying issues.  The first
> is that this system requires some model-specific handing in order to map
> chip outputs correctly.  The second is that if such model-specific details
> are required then the driver preferrably would activate these automatically
> without the need to specify "model=will".
>
> Another possibility is that there is something about this laptop system
> which is confusing the auto-probing of the alc260 codec code.
>
> I'm not sure what is the best way of proceeding here.  I suspect other more
> active developers will have a streamlined approach to help rectify the
> trouble.  In the first instance I would tend to load the module with
> "model=test" and experiment with the resulting alsamixer controls to
> determine what chip outputs are connected to which pins.  However, this is
> quite "old school" now - while it's the approach I used when getting my
> system working, my impression is that its use has been mostly deprecated by
> the automatic parser which is now present.
>
> In the above-referenced launchpad report the question is asked about the
> ALC260 datasheet.  This is readily available on the net.  I doubt it is all
> that necessary now since I think the base infrastructure is pretty solid,
> but if you do need it and have difficulty finding it please contact me
> off-list.
>

The auto-parser require correct pin default

Refer to alc260 General Description

Realtek's proprietary impedance sensing and jack detect techniques
allow device loads on inputs and outputs to be auto-detected. All
analog IOs are input and output capable. Headphone amplifiers are also
integrated at each analog output. All analog IOs can be re-tasked
according to user's definitions, or automatically switched to suit the
connecting device (Universal Audio Jack®).

Reserve analog mixer architecture is backwards compatible with AC'97