Add support for latency measurement to Telemetry smoothness benchmarks.

tools/perf/trace-info.json

 {
   "_description" : "This file contains info about our performance test traces, used by the perf dashboard (http://chromeperf.appspot.com). Links in test descriptions should be in markdown format. SUITE as the description means the same description as the test suite.",
 
   "dom": {
     "description": ""
   },
   "dom_attr": {
     "description": ""
   },
   "dom_attr_element_expando": {
@@ skipping 420 matching lines @@
   },
   "load_time": {
     "description": ""
   },
   "total_texture_upload_time": {
     "description": "The time spent in texture upload on the GPU process."
   },
   "load_time_by_url": {
     "description": ""
   },
+  "average_latency": {
+    "description": "The average latency between input event and corresponding swap to screen."
+  },
 
   "Score": {
     "description": "SUITE"
   },
   "canvasDrawImageFullClear": {
     "description": "Using a canvas element to render.  Bitmaps are blitted to the canvas using the 'drawImage' function and the canvas is fully cleared at the beginning of each frame."
   },
   "canvasDrawImageFullClearAlign": {
     "description": "Same as canvasDrawImageFullClear except all 'x' and 'y' values are roudned to the nearest integer.  This can be more efficient on translate on certain browsers."
   },
@@ skipping 88 matching lines @@
   "string-tagcloud": {
     "description": "String processing, including code to generate a giant 'tagcloud', extracting compressed JS code, etc."
   },
   "string-unpack-code": {
     "description": "String processing, including code to generate a giant 'tagcloud', extracting compressed JS code, etc."
   },
   "string-validate-input": {
     "description": "String processing, including code to generate a giant 'tagcloud', extracting compressed JS code, etc."
   }
 }