All graphs for this bug:
  https://chromeperf.appspot.com/group_report?bug_id=905183

(For debugging:) Original alerts at time of bug-filing:
  https://chromeperf.appspot.com/group_report?sid=746c23ccde1f1a744ff37a097cfb2ed534a2ec7d6eb3704221ce709ff48022fe


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

linux-perf

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

📍 Pinpoint job started.
https://pinpoint-dot-chromeperf.appspot.com/job/12aa702fe40000

📍 Found a significant difference after 1 commit.
https://pinpoint-dot-chromeperf.appspot.com/job/12aa702fe40000

Avoid unnecessary call to MakeCurrent by backer@chromium.org
https://chromium.googlesource.com/chromium/src/+/ce6ad323cd9aa0c49aa7245fe942dc5b9a91e421
tasks_per_frame_total_all: 378.4 → 444.2 (+65.78)

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

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

I was able to repro this locally. The reason the number of tasks has gone up is because we are checkerboarding less (rastering more pixels for display). The reason we are able to do this is because we are blocking the GPU main thread less and able to schedule more work onto it.