On Tue, 2013-08-06 at 11:00 -0700, Felipe Tonello wrote: > Hi Arun, > > On Mon, Aug 5, 2013 at 8:43 PM, Arun Raghavan > <arun.raghavan at collabora.co.uk> wrote: > > On Mon, 2013-08-05 at 11:43 -0700, Felipe Tonello wrote: > >> Hi all, > >> > >> I'm using UCM configurations to handle multiple devices and > >> playback/capture mixer controls. It seems that PA UCM implementation > >> just store the PlaybackVolume PlaybackSwitch properties but doesn't > >> make any usage of it. > >> > >> What I want to do is basically to be able to change hardware volume > >> instead of just software volume. When I plug a Headphone, for > >> instance, I want PA to change the volume of the alsa control that I > >> specified by PlaybackVolume in the UCM file not just the sink volume. > >> > >> Is that possible to do? Does it require a patch or some work? > > > > It is possible of course, but hasn't been implemented (yet). > > I will probably need this use case. I hope to send a patch in a few > weeks to come. Excellent! > > > >> Also, jack support for UCM seems to have a bug. For when using UCM > >> support, when I start PA or change a card profile with a jack > >> previously plugged, it doesn't recognizes that unless I unplug and > >> plug again. > > > > David was looking into hooking up jack detection with UCM, so might have > > some more useful information for you. > > Actually I realized was probably a problem in my UCM configuration. > Now it seems to be ok. > > > > >> Any comments on that as well? > >> > >> OBS: I'm using PA 3.0. > > > > You'd also mentioned the problem with modem PCMs on IRC. I think we > > really don't have much choice here but to go with the > > PA_SINK_ALWAYS_RUNNING flag that I'd posted earlier, coupled with adding > > a field in the UCM to signal that this needs to be set. > > I saw your patch. It was not upstreamed right? Also, how would you set > that flag in run-time? Correct about the former (though I'll likely rebase and resubmit it at some point soon). As for applying it at run time, we could add a value known to PA in the UCM 'Value' section, and then have our UCM processing set the flag based on that. > > > > Yes, we would ideally like for these paths to be set up in the kernel, > > but pragmatically, that seems to be quite unlikely to happen on the > > Android kernels (at least it hasn't in the last year and a half). > > Do you have more information about that? Because here I have control > over the kernel, so we could implement there if needed. Mark Brown (broonie on IRC) is probably the best person to answer this. Cheers, Arun
