[pulseaudio-discuss] [PATCH v2 0/6] ucm: Add support for "JackHWMute"

[Tanu Kaskinen]

On Mon, 2015-05-04 at 21:03 +0300, Tanu Kaskinen wrote:
> Here's v2 for supporting UCM's JackHWMute. What used to be only one
> patch is now six patches, to make the review experience even more
> pleasant. There are several changes beyond just splitting the
> patch into smaller pieces:
> 
>  - Instead of creating one jack object per UCM device name, there's
>    now one jack object per kcontrol. The reason for this is that I
>    gained motivation to refactor the jack handling in the traditional
>    mixer code too, and I plan to unify both code bases to use the one
>    jack per kcontrol model, which I think is the most conceptually
>    clear approach (creating one jack object per physical jack would be
>    even better, but Linux doesn't provide the necessary information to
>    implement that).
> 
>  - Removed input and output jack separation. UCM doesn't support
>    specifying separate jacks, and pretending otherwise made working
>    with the code more difficult than necessary.
> 
>  - JackHWMute can actually contain more than one device (use-case.h
>    documentation patch has been sent to alsa-devel). The code was
>    updated to reflect this.
> 
>  - UCM ports are now initialized with unknown availability to match
>    the traditional mixer code. This revealed a couple of bugs, which
>    are now fixed:
> 
>     - When the jack state was changed, the order of updating ordinary
>       and hw-mute devices' port availability was "wrong":
>       module-switch-on-port-available didn't like the original order.
>       Doing the updates in a different order made
>       module-switch-on-port-available happy. (There's a wall of text
>       comment in the code to explain the details.)
> 
>     - In v1, I did not take pa_alsa_jack.has_control into account when
>       updating port availability. That didn't seem to have much effect
>       when ports were initialized with "yes" availability, but with
>       "unknown" this bug became visible. Now has_control is always
>       considered when updating the port availability.
> 
>  - Added pa_alsa_jack_new() and pa_alsa_jack_free(). That's a minor
>    refactoring that I felt was useful.
> 
> Tanu Kaskinen (6):
>   alsa: Add pa_alsa_jack_new() and pa_alsa_jack_free()
>   ucm: Don't create separate input and output jacks
>   alsa: Add associations between jacks, UCM devices and UCM ports
>   alsa: Move UCM port availability updating to the mixer code
>   ucm: Create only one jack object per kcontrol
>   ucm: Add support for "JackHWMute"
> 
>  src/modules/alsa/alsa-mixer.c       | 138 ++++++++++++++++---
>  src/modules/alsa/alsa-mixer.h       |  10 ++
>  src/modules/alsa/alsa-ucm.c         | 268 +++++++++++++++++++++++++++++++-----
>  src/modules/alsa/alsa-ucm.h         |  16 ++-
>  src/modules/alsa/module-alsa-card.c |  19 +--
>  src/pulsecore/device-port.c         |  14 +-
>  6 files changed, 401 insertions(+), 64 deletions(-)

I agreed with David to push this patch series, even though the patches
haven't been through proper review. This patch series was the last thing
blocking a release candidate, so RC1 should happen soon.

-- 
Tanu