win-dshow: Actually fix LGP issue

The LGP issue is caused by the device drivers returning two or more
packets in a single segment of audio data.  This fixes it by detecting
that and decoding subsequent packets.

When the FFmpeg audio decoder returns, it returns how many bytes of data
was decoded.  To have it decode multiple packets in a single segment,
just subtract the return value from the expected size, and if that size
is still larger than zero, then there are more packets in the segment to
decode.  Otherwise, stop.

plugins/win-dshow/win-dshow.cpp

@@ -116,6 +116,7 @@ public:
 	inline ~Decoder() {ffmpeg_decode_free(&decode);}
 
 	inline operator ffmpeg_decode*() {return &decode;}
+	inline ffmpeg_decode *operator->() {return &decode;}
 };
 
 class CriticalSection {
@@ -509,21 +510,30 @@ void DShowInput::OnEncodedAudioData(enum AVCodecID id,
 		}
 	}
 
-	bool got_output;
-	int len = ffmpeg_decode_audio(audio_decoder, data, size,
-			&audio, &got_output);
-	if (len < 0) {
-		blog(LOG_WARNING, "Error decoding audio");
-		return;
-	}
+	do {
+		bool got_output;
+		int len = ffmpeg_decode_audio(audio_decoder, data, size,
+				&audio, &got_output);
+		if (len < 0) {
+			blog(LOG_WARNING, "Error decoding audio");
+			return;
+		}
 
-	if (got_output) {
-		audio.timestamp = (uint64_t)ts * 100;
+		if (got_output) {
+			audio.timestamp = (uint64_t)ts * 100;
 #if LOG_ENCODED_AUDIO_TS
-		blog(LOG_DEBUG, "audio ts: %llu", audio.timestamp);
+			blog(LOG_DEBUG, "audio ts: %llu", audio.timestamp);
 #endif
-		obs_source_output_audio(source, &audio);
-	}
+			obs_source_output_audio(source, &audio);
+		} else {
+			break;
+		}
+
+		ts += int64_t(audio_decoder->frame->nb_samples) * 10000000LL /
+			int64_t(audio_decoder->frame->sample_rate);
+		size -= (size_t)len;
+		data += len;
+	} while (size > 0);
 }
 
 void DShowInput::OnAudioData(const AudioConfig &config,