[telemetry] - Change default benchmark list to use chartjson output.

Waterfall is running on chartjson now, but we've got hacks in the recipes to swap the command lines from buildbot->chartjson.

ie. instead of:

{
  "steps": {
    "blink_perf.animation": {
      "cmd": "/path/to/run_benchmark -v --output-format=buildbot the_test",
      "device_affinity": -1,
      "perf_dashboard_id": "foo"
    },
}

->

{
  "steps": {
    "blink_perf.animation": {
      "cmd": "/path/to/run_benchmark -v --output-format=chartjson --upload-results the_test",
      "device_affinity": -1,
      "perf_dashboard_id": "foo"
    },
}

BUG=422173, 422174

Review URL: https://codereview.chromium.org/779223004

Cr-Commit-Position: refs/heads/master@{#307072}

1 files changed, 2 insertions, 1 deletions

diff --git a/tools/telemetry/telemetry/benchmark_runner.py b/tools/telemetry/telemetry/benchmark_runner.py
index 27b3c27..0c0a924 100644
--- a/tools/telemetry/telemetry/benchmark_runner.py
+++ b/tools/telemetry/telemetry/benchmark_runner.py
@@ -245,7 +245,8 @@ def _GetJsonBenchmarkList(possible_browser, possible_reference_browser,
 
     base_name = benchmark_class.Name()
     base_cmd = [sys.executable, os.path.realpath(sys.argv[0]),
-                '-v', '--output-format=buildbot', base_name]
+                '-v', '--output-format=chartjson', '--upload-results',
+                base_name]
     perf_dashboard_id = base_name
     # TODO(tonyg): Currently we set the device affinity to a stable hash of the
     # benchmark name. This somewhat evenly distributes benchmarks among the