All graphs for this bug:
  https://chromeperf.appspot.com/group_report?bug_id=895306

(For debugging:) Original alerts at time of bug-filing:
  https://chromeperf.appspot.com/group_report?sid=aead7e2e93aa811ef79429f763526ed48f391dbd0bf40c120a3c24ff4d2d148f


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

Android Nexus5X WebView Perf

memory.top_10_mobile - Benchmark documentation link:
  None

📍 Pinpoint job started.
https://pinpoint-dot-chromeperf.appspot.com/job/13d18632e40000

📍 Found a significant difference after 1 commit.
https://pinpoint-dot-chromeperf.appspot.com/job/13d18632e40000

[async] Introduce dedicated JSAsyncFunctionObject. by bmeurer@chromium.org
https://chromium.googlesource.com/v8/v8/+/a63987a41aa9d0bc6ba120376ca7511f4196624b
memory:webview:all_processes:reported_by_os:system_memory:native_heap:proportional_resident_size: 3.232e+07 → 3.253e+07 (+2.153e+05)

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

Benchmark documentation link:
  None

This moves the promise from the scope (aka bytecode register file) to the async function object. I doubt that this site uses async functions at all, so this result seems wrong.