All graphs for this bug:
  https://chromeperf.appspot.com/group_report?bug_id=667684

Original alerts at time of bug-filing:
  https://chromeperf.appspot.com/group_report?keys=agxzfmNocm9tZXBlcmZyFAsSB0Fub21hbHkYgICgq6vfzAkM,agxzfmNocm9tZXBlcmZyFAsSB0Fub21hbHkYgICg6-HKowoM,agxzfmNocm9tZXBlcmZyFAsSB0Fub21hbHkYgICg64WZvwoM,agxzfmNocm9tZXBlcmZyFAsSB0Fub21hbHkYgICg68GzrgsM,agxzfmNocm9tZXBlcmZyFAsSB0Fub21hbHkYgICg64XStwoM,agxzfmNocm9tZXBlcmZyFAsSB0Fub21hbHkYgICgq7Lu5QgM


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

webrtc-win-large-tests

magjed@: These regressions seems possibly to be related either to your CL https://codereview.webrtc.org/2493133002 or to the CL https://codereview.webrtc.org/2505183003 which changed the DCHECK behavior on the perf-bots. Do you have any idea of whether either of them could have caused this regression?

I think it's very unlikely that it's my CL. The DCHECK CL sounds much more plausible. kjellander@ can you confirm it's your CL?

When removing DCHECKs we expected improvements alerts, as I e-mailed the perf sheriffs about, so yes.