Update a ton of audio code to use size_t more correctly and in general reduce

webrtc/tools/agc/test_utils.cc

Index: webrtc/tools/agc/test_utils.cc
diff --git a/webrtc/tools/agc/test_utils.cc b/webrtc/tools/agc/test_utils.cc
index 3a26cb9ac79a45fc39ac969f33f3ec50f40e9b98..81819c598e38aa7784a5eac20e7ab544506f007d 100644
--- a/webrtc/tools/agc/test_utils.cc
+++ b/webrtc/tools/agc/test_utils.cc
@@ -27,11 +27,12 @@ float Db2Linear(float db) {
 }
 
 void ApplyGainLinear(float gain, float last_gain, AudioFrame* frame) {
-  const int frame_length = frame->samples_per_channel_ * frame->num_channels_;
+  const size_t frame_length =
+      frame->samples_per_channel_ * frame->num_channels_;
   // Smooth the transition between gain levels across the frame.
   float smoothed_gain = last_gain;
   float gain_step = (gain - last_gain) / (frame_length - 1);
-  for (int i = 0; i < frame_length; ++i) {
+  for (size_t i = 0; i < frame_length; ++i) {
     smoothed_gain += gain_step;
     float sample = std::floor(frame->data_[i] * smoothed_gain + 0.5);
     sample = std::max(std::min(32767.0f, sample), -32768.0f);