All graphs for this bug:
  https://chromeperf.appspot.com/group_report?bug_id=898906

(For debugging:) Original alerts at time of bug-filing:
  https://chromeperf.appspot.com/group_report?sid=b0a5e1ca8c33e5cc7a1ba5bc9f2a0089137cac44cd15326e1e8c247a2b9266b6


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

Android Nexus5X WebView Perf
Android Nexus6 WebView Perf
android-nexus5x-perf
mac-10_12_laptop_low_end-perf

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

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

memory.top_10_mobile - Benchmark documentation link:
  None

📍 Pinpoint job started.
https://pinpoint-dot-chromeperf.appspot.com/job/171076eee40000

📍 Found a significant difference after 1 commit.
https://pinpoint-dot-chromeperf.appspot.com/job/171076eee40000

gpu: Fix raster implementation error for no flush result available query by khushalsagar@chromium.org
https://chromium.googlesource.com/chromium/src/+/da82bcce69bf5384435976a1945c9cb3270c5abb
memory:chrome:all_processes:reported_by_chrome:gpu:effective_size: 4096 → 1.577e+06 (+1.573e+06)

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

Benchmark documentation link:
  None

All graphs have recovered with the change for rate limiting the query (https://chromium-review.googlesource.com/c/1292110)

I'm going to guess this was another case of these queries changing the timing of cleanup when combined with the change above for removing the accidental flush that was added with the change which introduced these raster queries. Either way, its fixed now.