google_rtc_audio_processing: Handle inactive reference input

Close the loop on "optional pipeline" state management.  When the
component connected to the reference input is inactive, synthesize
zeros in the buffer being fed to echo cancellation processing.

Now the capture pipe is no longer necessary and no host DMA needs to
be happening on the playback channel just to feed AEC.  You can start
and stop playback at any time and the stream will accommodate.

Note that mic startup happens on the host's schedule and isn't
synchronous, so there's a case when exercising this code where the
reference stream can be a few 10ms blocks ahead of the mic when it
finally arrives.  Add a "catch up" check to discard stale samples.

Signed-off-by: Andy Ross <[email protected]>

src/audio/google/google_rtc_audio_processing.c

@@ -713,23 +713,45 @@ static int google_rtc_audio_processing_process(struct processing_module *mod,
 	struct audio_stream *ref = input_buffers[cd->aec_reference_source].data;
 	struct audio_stream *out = output_buffers[0].data;
 
+	struct comp_buffer *refbuf = container_of(ref, struct comp_buffer, stream);
+	bool ref_ok = refbuf->source->state == COMP_STATE_ACTIVE;
+
+	/* Would be cleaner to store a bit of state to elide a bzero
+	 * we already did, but we'd be doing the copy of real data in
+	 * the ref_ok state anyway.
+	 */
+	if (!ref_ok)
+		bzero(cd->aec_reference_buffer,
+		      (cd->num_frames * cd->num_aec_reference_channels
+		       * sizeof(cd->aec_reference_buffer[0])));
+
 	int micchan = audio_stream_get_channels(mic);
 	int refchan = audio_stream_get_channels(ref);
 
 	int fmic = audio_stream_get_avail_frames(mic);
 	int fref = audio_stream_get_avail_frames(ref);
-	int frames = MIN(fmic, fref);
+	int frames = ref_ok ? MIN(fmic, fref) : fmic;
 	int n, frames_rem;
 
+	/* If fref > fmic (common at pipeline startup if
+	 * playback was already active), we should consume the early
+	 * samples so AEC compares the most recent values.
+	 */
+	if (ref_ok && fref > fmic)
+		audio_stream_consume(ref, (fref - fmic) * audio_stream_frame_bytes(ref));
+
 	for (frames_rem = frames; frames_rem; frames_rem -= n) {
 		n = MIN(frames, cd->num_frames - cd->raw_mic_buffer_frame_index);
 
 		audio_stream_copy_to_linear(mic, 0, cd->raw_mic_buffer,
 					    cd->raw_mic_buffer_frame_index,
 					    n * micchan);
-		audio_stream_copy_to_linear(ref, 0, cd->aec_reference_buffer,
-					    cd->aec_reference_frame_index,
-					    n * refchan);
+
+		if (ref_ok)
+			audio_stream_copy_to_linear(ref, 0, cd->aec_reference_buffer,
+						    cd->aec_reference_frame_index,
+						    n * refchan);
+
 		cd->raw_mic_buffer_frame_index += n;
 
 		if (cd->raw_mic_buffer_frame_index >= cd->num_frames) {