Major AEC3 render pipeline changes

webrtc/modules/audio_processing/test/aec_dump_based_simulator.cc

Index: webrtc/modules/audio_processing/test/aec_dump_based_simulator.cc
diff --git a/webrtc/modules/audio_processing/test/aec_dump_based_simulator.cc b/webrtc/modules/audio_processing/test/aec_dump_based_simulator.cc
index f444967a108955069df3002f0c934e108b83e09a..d1cd48424a1cfcb3708c170f2a1a91d866f35887 100644
--- a/webrtc/modules/audio_processing/test/aec_dump_based_simulator.cc
+++ b/webrtc/modules/audio_processing/test/aec_dump_based_simulator.cc
@@ -93,7 +93,7 @@ void AecDumpBasedSimulator::PrepareProcessStreamCall(
                  static_cast<size_t>(msg.input_channel_size()));
 
     // Populate input buffer.
-    for (int i = 0; i < msg.input_channel_size(); ++i) {
+    for (size_t i = 0; i < in_buf_->num_channels(); ++i) {
       RTC_CHECK_EQ(in_buf_->num_frames() * sizeof(*in_buf_->channels()[i]),
                    msg.input_channel(i).size());
       std::memcpy(in_buf_->channels()[i], msg.input_channel(i).data(),