All graphs for this bug:
  https://chromeperf.appspot.com/group_report?bug_id=918539

(For debugging:) Original alerts at time of bug-filing:
  https://chromeperf.appspot.com/group_report?sid=fbd7a08e594308b03aba32df7e65576ef28b936227fe7da58ac1dd0202927b6f


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

Win 7 Nvidia GPU Perf
Win 7 Perf
mac-10_12_laptop_low_end-perf

rendering.desktop - Benchmark documentation link:
  https://bit.ly/rendering-benchmarks

📍 Pinpoint job started.
https://pinpoint-dot-chromeperf.appspot.com/job/11a26f44940000

📍 Found significant differences after each of 2 commits.
https://pinpoint-dot-chromeperf.appspot.com/job/11a26f44940000

Revert "ARM64 capable toolchain hash" by machenbach@chromium.org
https://chromium.googlesource.com/chromium/src/+/ee86e3d1d70c04102da0a2311f90819f343aee8e
frame_times: No values → 123.9

Reland "Reland "Disable the new tab-loading animation"" by pbos@chromium.org
https://chromium.googlesource.com/chromium/src/+/831dad6e5036b31b683587078db3df8e32ce4f92
frame_times: 147.1 → 743.7 (+596.6)

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

Benchmark documentation link:
  https://bit.ly/rendering-benchmarks

My regression was due to reverting an experiment that never shipped, so it shouldn't have been a "real" improvement.

It likely made more frames of the tab-icon animation play on a layer which might've lowered the total number of large-texture uploads.