[scheduler] Monitor renderer load level.

tools/metrics/histograms/histograms.xml

Index: tools/metrics/histograms/histograms.xml
diff --git a/tools/metrics/histograms/histograms.xml b/tools/metrics/histograms/histograms.xml
index 75f16fdd33e2dc9286685a4235553e439a553fba..ed9c67cafa7554870cac3e6fba32f2afd7182df3 100644
--- a/tools/metrics/histograms/histograms.xml
+++ b/tools/metrics/histograms/histograms.xml
@@ -47574,6 +47574,14 @@ http://cs/file:chrome/histograms.xml - but prefer this file for new entries.
   </summary>
 </histogram>
 
+<histogram name="RendererScheduler.BackgroundRendererLoad" units="%">
+  <owner>altimin@chromium.org</owner>
+  <summary>
+    Background renderer load level i.e. percentage of the time spent on running
+    tasks.

[Mark P, 2016/08/22 17:28:38]

Every histogram should mention when it is emitted.

Also, it would be good to clarity how this is affected by number of cores. 
Usually, say, a 4-core machine can go up to a load of 4, yet this description of
yours implies it can only go up to a "load" of one because you measure load as
percent of time.  Which is it?  (Depending on your answer, you may also want to
revise the histogram name.)

[altimin, 2016/08/22 17:39:33]

Actually, it's main thread load. Changed name to reflect this.

+  </summary>
+</histogram>
+
 <histogram name="RendererScheduler.ExpectedTaskQueueingDuration" units="ms">
   <owner>tdresser@chromium.org</owner>
   <summary>
@@ -47583,6 +47591,14 @@ http://cs/file:chrome/histograms.xml - but prefer this file for new entries.
   </summary>
 </histogram>
 
+<histogram name="RendererScheduler.ForegroundRendererLoad" units="%">
+  <owner>altimin@chromium.org</owner>
+  <summary>
+    Foreground renderer load level i.e. percentage of the time spent on running
+    tasks.
+  </summary>
+</histogram>
+
 <histogram name="RendererScheduler.TaskQueueManager.DelayedTaskLateness"
     units="ms">
   <owner>alexclarke@chromium.org</owner>