Support needed to measure user and service traffic in Chrome.

chrome/browser/net/chrome_network_delegate.cc

Index: chrome/browser/net/chrome_network_delegate.cc
diff --git a/chrome/browser/net/chrome_network_delegate.cc b/chrome/browser/net/chrome_network_delegate.cc
index bed47f796f2a1f8d21426ade117ada47c08e8877..b62274e15fee87c23aef5e0bca1b0e6fecc72fb5 100644
--- a/chrome/browser/net/chrome_network_delegate.cc
+++ b/chrome/browser/net/chrome_network_delegate.cc
@@ -491,6 +491,13 @@ void ChromeNetworkDelegate::OnNetworkBytesReceived(
 
 void ChromeNetworkDelegate::OnCompleted(net::URLRequest* request,
                                         bool started) {
+#if !defined(OS_IOS)
+  // TODO(amohammadkhan): Reporting should be done on redirects too. Because
+  // following a redirect resets the received/sent bytes. Though if the request
+  // is cancelled instead of following the redirect, it will end up here, and
+  // the data will be double counted.
+  data_use_measurement_.ReportDataUseUMA(request);

[mmenke, 2015/09/03 16:02:58]

Think first pass, you should measure usage for redirects.  Can worry about the
duplicate data later (If at all).

[amohammadkhan, 2015/09/03 23:10:36]

Sorry I think I didn't understand what you meant first time. Now I added the
code for redirects and a TODO for double counted.

+#endif
   RecordNetworkErrorHistograms(request);
   if (started) {
     // Only call in for requests that were started, to obey the precondition