All graphs for this bug:
  https://chromeperf.appspot.com/group_report?bug_id=600489

Original alerts at time of bug-filing:
  https://chromeperf.appspot.com/group_report?keys=agxzfmNocm9tZXBlcmZyFAsSB0Fub21hbHkYgICgyO_ItQoM,agxzfmNocm9tZXBlcmZyFAsSB0Fub21hbHkYgICgqIjzvgoM,agxzfmNocm9tZXBlcmZyFAsSB0Fub21hbHkYgICgyLmhsQoM


Bot(s) for this bug's original alert(s):

webrtc-linux-large-tests

r12204 == https://codereview.webrtc.org/1855003002, which unfortunately is a relatively large roll.

However, this may be a dupe of  bug 600486  because I do see the signs of an increase starting with r12203. Assigning to this week's perf sheriff to follow up.

This is the expected behavior, undoing accidental "improvements" (really just timing changes) introduced in  bug 597332 . If you zoom out the chart, you'll see they just revert to levels from before the bug.