Adding APM configuration in AEC dump.

webrtc/modules/audio_processing/debug.proto

Index: webrtc/modules/audio_processing/debug.proto
diff --git a/webrtc/modules/audio_processing/debug.proto b/webrtc/modules/audio_processing/debug.proto
index dce2f7920930452f7a512d9bda2f3679893cddc2..520eac94c80616902e6650bdc7388803ba8fa3bf 100644
--- a/webrtc/modules/audio_processing/debug.proto
+++ b/webrtc/modules/audio_processing/debug.proto
@@ -39,11 +39,36 @@ message Stream {
   repeated bytes output_channel = 8;
 }
 
+message Config {
+  // Acoustic echo canceler
+  optional bool aec_enabled = 1;
+  optional bool aec_delay_agnostic = 2;
+  optional bool aec_drift_compensation = 3;
+  optional bool aec_extended_filter = 4;
+  optional int32 aec_suppression_level = 5;
+  // Mobile AEC
+  optional bool aecm_enabled = 11;

[ivoc, 2015/09/25 09:24:21]

Shouldn't we continue with field number 6 here? Or are you reserving space for
future fields?

[minyue-webrtc, 2015/09/25 09:49:09]

Yes, precisely. I feel that there can be info to add in future, and giving some
gaps will help readability for future. Ivo, are there any disadvantages that I
may not be aware of?

[ivoc, 2015/09/25 10:22:35]

Well, fields 1 to 15 take a single byte to encode, while fields 16 to 2047 take
2 bytes to encode. However, I think this is not that important for this message,
since there will not be that many Config messages in a typical log file. 

[Andrew MacDonald, 2015/09/26 02:36:19]

Personally I wouldn't bother with this. You can keep a "next field number"
comment at the top, so people don't need to scan the list to find it.

[minyue-webrtc, 2015/09/26 16:26:44]

I think that is nice.

+  optional bool aecm_comfort_noise = 12;
+  optional int32 aecm_routing_mode = 13;
+  // Automatic gain controller
+  optional bool agc_enabled = 21;
+  optional bool agc_experiment = 22;
+  optional int32 agc_mode = 23;
+  optional bool agc_limiter = 24;
+  // High pass filter
+  optional bool hpf_enabled = 31;
+  // Noise suppression
+  optional bool ns_enabled = 41;
+  optional bool ns_experiment = 42;
+  optional int32 ns_level = 43;
+}
+
 message Event {
   enum Type {
     INIT = 0;
     REVERSE_STREAM = 1;
     STREAM = 2;
+    CONFIG = 4;

[ivoc, 2015/09/25 09:24:21]

Shouldn't this be value 3?

[minyue-webrtc, 2015/09/25 09:49:09]

Yes indeed

   }
 
   required Type type = 1;
@@ -51,4 +76,5 @@ message Event {
   optional Init init = 2;
   optional ReverseStream reverse_stream = 3;
   optional Stream stream = 4;
+  optional Config config = 5;
 }