[pulseaudio-discuss] [PATCH] sink: Make early drain reporting propagate down to filters

[Alexander E. Patrakov]

17.11.2015 12:13, arun at accosted.net wrote:
> From: Arun Raghavan <git at arunraghavan.net>
>
> The drain reporting improvements that were added to alsa-sink were only
> being applied to directly connected sink inputs. This patch makes the
> same logic also recurse down the filter hierarchy, so drains are
> acknowledged more accurately (and not late) even if there is a filter
> sink in between.
>
> Also does some minor reorganisation of the code and sprinkles in some
> comments as documentation.

Thanks for the patch. Do I understand correctly that it supersedes this one?

http://thread.gmane.org/gmane.comp.audio.pulseaudio.general/21582

-- 
Alexander E. Patrakov

> ---
>   src/pulsecore/sink.c | 38 +++++++++++++++++++++++++++++++-------
>   1 file changed, 31 insertions(+), 7 deletions(-)
>
> diff --git a/src/pulsecore/sink.c b/src/pulsecore/sink.c
> index 9ddb527..0b44fc7 100644
> --- a/src/pulsecore/sink.c
> +++ b/src/pulsecore/sink.c
> @@ -950,18 +950,42 @@ size_t pa_sink_process_input_underruns(pa_sink *s, size_t left_to_play) {
>
>       PA_HASHMAP_FOREACH(i, s->thread_info.inputs, state) {
>           size_t uf = i->thread_info.underrun_for_sink;
> -        if (uf == 0)
> -            continue;
> -        if (uf >= left_to_play) {
> -            if (pa_sink_input_process_underrun(i))
> -                continue;
> +
> +        /* Propagate down the filter tree */
> +        if (i->origin_sink) {
> +            size_t filter_result, left_to_play_origin;
> +
> +            /* The recursive call works in the origin sink domain ... */
> +            left_to_play_origin = pa_convert_size(left_to_play, &i->sink->sample_spec, &i->origin_sink->sample_spec);
> +
> +            /* .. and returns the time to sleep before waking up. We need the
> +             * underrun duration for comparisons, so we undo the subtraction on
> +             * the return value... */
> +            filter_result = left_to_play_origin - pa_sink_process_input_underruns(i->origin_sink, left_to_play_origin);
> +
> +            /* ... and convert it back to the master sink domain */
> +            filter_result = pa_convert_size(filter_result, &i->origin_sink->sample_spec, &i->sink->sample_spec);
> +
> +            /* Remember the longest underrun so far */
> +            if (filter_result > result)
> +                result = filter_result;
>           }
> -        else if (uf > result)
> +
> +        if (uf == 0) {
> +            /* No underrun here, move on */
> +            continue;
> +        } else if (uf >= left_to_play) {
> +            /* The sink has possibly consumed all the data the sink input provided */
> +            pa_sink_input_process_underrun(i);
> +        } else if (uf > result) {
> +            /* Remember the longest underrun so far */
>               result = uf;
> +        }
>       }
>
>       if (result > 0)
> -        pa_log_debug("Found underrun %ld bytes ago (%ld bytes ahead in playback buffer)", (long) result, (long) left_to_play - result);
> +        pa_log_debug("%s: Found underrun %ld bytes ago (%ld bytes ahead in playback buffer)", s->name,
> +                (long) result, (long) left_to_play - result);
>       return left_to_play - result;
>   }
>
>


-- 
Alexander E. Patrakov