GitLab

The image decoding benchmark still contained some logic for navigation/test
control. The image_decoding_benchmark should measure image decodign times. Critically, the
logic to make chrome/test/data/image_decoding/image_decoding.html do its work
should be handled by the page runner, not the benchmark. To make this
separation happen, three changes are made...

First, page runner now notifies the test that naivgation will happen. This
allows us to begin recording the timeline right before navigation. This is when timeline
recording should begin for image decode tests.

Second, this adds a post_navigate_javascript_to_execute method to the page. This allows us to
call runBenchmark() on the image decoding benchmarks after we navigate.

Finally, the benchmark uses the standard wait_for_javascript_expression to poll
the isDone variable.The old code actually did this in python.

R=tonyg
NOTRY=True

Review URL: https://chromiumcodereview.appspot.com/11753023

git-svn-id: svn://svn.chromium.org/chrome/trunk/src@175404 0039d316-1c4b-4281-b951-d872f2087c98