audio: make decoders output refcounted frames

This rewrites the audio decode loop to some degree. Audio filters don't
do refcounted frames yet, so af.c contains a hacky "emulation".

Remove some of the weird heuristic-heavy code in dec_audio.c. Instead of
estimating how much audio we need to filter, we always filter full
frames. Maybe this should be adjusted later: in case filtering increases
the volume of the audio data, we should try not to buffer too much
filter output by reducing the input that is fed at once.

For ad_spdif.c and ad_mpg123.c, we don't avoid extra copying yet - it
doesn't seem worth the trouble.

2 files changed, 8 insertions, 8 deletions

diff --git a/player/audio.c b/player/audio.c
index 413cbf4a32..e3dbe6577c 100644
--- a/player/audio.c
+++ b/player/audio.c
@@ -192,6 +192,7 @@ void reinit_audio_chain(struct MPContext *mpctx)
         mpctx->d_audio->global = mpctx->global;
         mpctx->d_audio->opts = opts;
         mpctx->d_audio->header = sh;
+        mpctx->d_audio->pool = mp_audio_pool_create(mpctx->d_audio);
         mpctx->d_audio->afilter = af_new(mpctx->global);
         mpctx->d_audio->afilter->replaygain_data = sh->audio->replaygain_data;
         mpctx->ao_buffer = mp_audio_buffer_create(NULL);
@@ -207,8 +208,7 @@ void reinit_audio_chain(struct MPContext *mpctx)
     }
     assert(mpctx->d_audio);
 
-    struct mp_audio in_format;
-    mp_audio_buffer_get_format(mpctx->d_audio->decode_buffer, &in_format);
+    struct mp_audio in_format = mpctx->d_audio->decode_format;
 
     if (!mp_audio_config_valid(&in_format)) {
         // We don't know the audio format yet - so configure it later as we're
@@ -238,7 +238,7 @@ void reinit_audio_chain(struct MPContext *mpctx)
     }
 
     // filter input format: same as codec's output format:
-    mp_audio_buffer_get_format(mpctx->d_audio->decode_buffer, &afs->input);
+    afs->input = in_format;
 
     // Determine what the filter chain outputs. recreate_audio_filters() also
     // needs this for testing whether playback speed is changed by resampling
@@ -304,8 +304,7 @@ double written_audio_pts(struct MPContext *mpctx)
     if (!d_audio)
         return MP_NOPTS_VALUE;
 
-    struct mp_audio in_format;
-    mp_audio_buffer_get_format(d_audio->decode_buffer, &in_format);
+    struct mp_audio in_format = d_audio->decode_format;
 
     if (!mp_audio_config_valid(&in_format) || d_audio->afilter->initialized < 1)
         return MP_NOPTS_VALUE;
@@ -324,7 +323,8 @@ double written_audio_pts(struct MPContext *mpctx)
     // Subtract data in buffers between decoder and audio out.
 
     // Decoded but not filtered
-    a_pts -= mp_audio_buffer_seconds(d_audio->decode_buffer);
+    if (d_audio->waiting)
+        a_pts -= d_audio->waiting->samples / (double)in_format.rate;
 
     // Data buffered in audio filters, measured in seconds of "missing" output
     double buffered_output = af_calc_delay(d_audio->afilter);
diff --git a/player/command.c b/player/command.c
index 04c0c4d789..3513d99afc 100644
--- a/player/command.c
+++ b/player/command.c
@@ -1550,7 +1550,7 @@ static int mp_property_samplerate(void *ctx, struct m_property *prop,
     MPContext *mpctx = ctx;
     struct mp_audio fmt = {0};
     if (mpctx->d_audio)
-        mp_audio_buffer_get_format(mpctx->d_audio->decode_buffer, &fmt);
+        fmt = mpctx->d_audio->decode_format;
     if (!fmt.rate)
         return M_PROPERTY_UNAVAILABLE;
     if (action == M_PROPERTY_PRINT) {
@@ -1567,7 +1567,7 @@ static int mp_property_channels(void *ctx, struct m_property *prop,
     MPContext *mpctx = ctx;
     struct mp_audio fmt = {0};
     if (mpctx->d_audio)
-        mp_audio_buffer_get_format(mpctx->d_audio->decode_buffer, &fmt);
+        fmt = mpctx->d_audio->decode_format;
     if (!fmt.channels.num)
         return M_PROPERTY_UNAVAILABLE;
     switch (action) {