All graphs for this bug:
  https://chromeperf.appspot.com/group_report?bug_id=825819

(For debugging:) Original alerts at time of bug-filing:
  https://chromeperf.appspot.com/group_report?sid=a95c9c22ab4fa8c0844b6e296aaa400fba6b5ffbd43258121cdc284c724ed2f3


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

android-webview-nexus6

📍 Pinpoint job started.
https://pinpoint-dot-chromeperf.appspot.com/job/11fa07a3440000

📍 Found significant differences after each of 4 commits.
https://pinpoint-dot-chromeperf.appspot.com/job/11fa07a3440000

DevTools: Move record allocation stacks option to the memory panel. by alph@chromium.org
https://chromium.googlesource.com/chromium/src/+/8b97320b87d459af6f3a5bad41f0de1681da394b

Increase hard timeout for win_chromium_dbg_ng due to some long-running tests. by jchinlee@chromium.org
https://chromium.googlesource.com/chromium/src/+/8478ca4049b9e6863da8461005b49acc4adb6803

Remove GetCommittedKBytes and GetWorkingSetKBytes. by erikchen@chromium.org
https://chromium.googlesource.com/chromium/src/+/52b7b6b65c81f214f3dfe0324d33e02281e7bced

[Feed] Updating the builds files for third_party/feed by fgorski@google.com
https://chromium.googlesource.com/chromium/src/+/b27239cfb1c2855a831a4107cbed4f0e82ab9a9e

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

📍 Pinpoint job started.
https://pinpoint-dot-chromeperf.appspot.com/job/16fa764bc40000

📍 Found a significant difference after 1 commit.
https://pinpoint-dot-chromeperf.appspot.com/job/16fa764bc40000

Change WebContentsObserver::SubresourceResponseStarted to load complete by jcivelli@google.com
https://chromium.googlesource.com/chromium/src/+/d5904f060dea9d9cecb9afb378a68a17fd4e8e25

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

jcivelli - Per comment #5, this regression is being caused by your CL, PTAL

Issue 825633 has been merged into this issue.

jcivelli please do try to reproduce the issue again, now you might find https://pinpoint-dot-chromeperf.appspot.com/ useful to do so.

Note the regression is pretty clear in java_heap:allocated_objects_size during infinite_scroll stories.