[alsa-devel] idt blue jack patch

[Takashi Iwai]

At Tue, 17 Feb 2009 23:04:31 -0300,
Paulo Cavalcanti wrote:
> 
> On Mon, Feb 16, 2009 at 7:55 AM, Takashi Iwai <tiwai at suse.de> wrote:
> 
>     At Mon, 16 Feb 2009 06:13:35 -0300,
>     Paulo Cavalcanti wrote:
>     >
>     > Hi,
>     >
>     > I received a patch form Fengguang that
>     > made my Intel DG45ID (snd-hda-intel, codec IDT 92HD73E1X5)
>     > behave the way it should, in my opinion.
>     >
>     > It just makes the blue jack behave as an input source by default (in
>     alsa
>     > 1.0.19 it is
>     > an output jack).
>     >
>     > It is widespread that the green jack is output, the blue input and the
>     pink
>     > is mic.
>     > Some cards allow reconfiguring the jacks in alsamixer, but the DG45ID
>     > already has
>     > an orange and black jacks for rear and center speakers, and a digital
>     > output.
>     >
>     > Other people have also raised this problem some time ago:
>     >
>     > https://fcp.surfsite.org/modules/newbb/viewtopic.php?viewmode=threaded&
>     order=ASC&topic_id=62855&forum=12&move=prev&topic_time=1224179600
>     >
>     > Fengguang suggested me to direct my request to the development list,
>     > so the other developers could decide what to do.
>     >
>     > I would like to add that I can use the input source and the mic just
>     fine
>     > now
>     > (the mic could be a bit louder, but it works fine with skype).
>     >
>     > The patch is very simple and is attached. This would conclude the fix
>     > for this card, Takashi started with me some time ago (I just made the
>     > tests).
>    
>     It's fine to put a quirk there.  But, you'd need to create a "model"
>     for your device so that it won't interfere other devices.  That is,
>     define a new model type, and call stac92xx_set_config_reg() with the
>     check of spec->board_config.
> 
> This is easy. The problem is that I am already using model "no-jd", remember?
> In this case, if I used
> 
>         /* was 0x01113014 */
>         if (spec->board_config > STAC_92HD73XX_NO_JD ) {
>             stac92xx_set_config_reg(codec, 0x0c, 0x01813021);
>         }
> 
> would be fine to me.

Hm, now the question is why it's wrong.  This patch changes the
port-C.  My concern is whether it's a cabling problem or if it's a
BIOS thing.  If it's a cabling problem (i.e. the cables don't match
with the box), it's not a problem that the driver needs to fix in it.
OTOH, if the cabling is OK but just the port-C is mislabeled by BIOS,
it's an issue to be fixed as a quirk in the driver.

> However, there is a possibility that all "stac92hd73xx" family needs
> this quirk. In this case, no additional code is needed.

Very unlikely.
The pins can be freely assigned, and the port-C is a very basic port.
The position, color, association, etc must be different from machine
to machine.


thanks,

Takashi