[pulseaudio-discuss] [WIP] Passthrough support

[Kelly Anderson]

On 03/16/11 10:48, Arun Raghavan wrote:
> On Wed, 2011-03-16 at 11:19 +0530, Arun Raghavan wrote:
> [...]
>> It will work, but not with paplay, only with the extended API when you
>> create a non-PCM stream. I will rationalise the check for passthrough
>> sink inputs in a bit, but for now, could you change the checks (there
> I've pushed a more comprehensive fix, so you shouldn't need to change
> that line and using paplay should work for you.
>
> -- Arun
>
Arun,

Hey, I just got around to testing without the following patch.  It 
doesn't work.  I also didn't see any recent checkins in the passthrough 
tree that would seem to affect this.  Is it possible the fix didn't get 
pushed?

--- ./src/modules/alsa/alsa-sink.c.orig 2011-03-16 01:17:02.061953124 -0600
+++ ./src/modules/alsa/alsa-sink.c      2011-03-16 01:19:07.492460278 -0600
@@ -1058,7 +1058,7 @@ static int sink_process_msg(pa_msgobject
              pa_sink_input *i = PA_SINK_INPUT(data);
              int r = 0;

-            if (PA_LIKELY(pa_format_info_is_pcm(i->format)))
+            if (PA_LIKELY(!(i->flags & PA_SINK_INPUT_PASSTHROUGH)) )
                  break;

              u->old_rate = u->sink->sample_spec.rate;
@@ -1084,7 +1084,7 @@ static int sink_process_msg(pa_msgobject
              pa_sink_input *i = PA_SINK_INPUT(data);
              int r = 0;

-            if (PA_LIKELY(pa_format_info_is_pcm(i->format)))
+            if (PA_LIKELY(!(i->flags & PA_SINK_INPUT_PASSTHROUGH)) )
                  break;

              /* Passthrough format, see if we need to reset sink sample 
rate */