All graphs for this bug:
  https://chromeperf.appspot.com/group_report?bug_id=880314

(For debugging:) Original alerts at time of bug-filing:
  https://chromeperf.appspot.com/group_report?sid=5d79f60ef90eb4cee9283cb9a67f76ce96c744dcd0d03a32c06bfb0302914a28


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

linux-perf
mac-10_12_laptop_low_end-perf
mac-10_13_laptop_high_end-perf

system_health.common_desktop - Benchmark documentation link:
  https://bit.ly/system-health-benchmarks

📍 Pinpoint job started.
https://pinpoint-dot-chromeperf.appspot.com/job/1798c205640000

📍 Found a significant difference after 1 commit.
https://pinpoint-dot-chromeperf.appspot.com/job/1798c205640000

Revert "Reland "Enable Perfetto by default for all telemetry tests"" by oysteine@chromium.org
https://chromium.googlesource.com/chromium/src/+/00bea4ec5efd575a09604b89fbf6123b14690fb6
0.04428 → 0.07216 (+0.02788)

Understanding performance regressions:
  http://g.co/ChromePerformanceRegressions

Reverted in https://chromium-review.googlesource.com/c/chromium/src/+/1194837

Oh nevermind, the regression was caused by the revert and not the original CL..? That's unexpected.

Oystein, did the original change cause improvement so the reverting caused regression?

Yep, and looks like the graphs all improved again with crrev.com/603931

Generally these are very low cpu_time_percentage_avg values, so changes in measurement overhead are pretty noticeable.