No idea what to do with these failures, unless I'm missing something the bot output doesn't give any information whatsoever other than "there is a leak"...

This is the output:

[ RUN      ] memory.leak_detection/https://www.tumblr.com/
  (WARNING) 2018-10-04 23:41:39,036 inspector_log._OnMessage:26  DevTools console [network]: Failed to load resource: the server responded with a status of 404 (Not Found) https://i2-zzzzzzzzzzzzzzzzzbbbbbbbbbbbbb.init.cedexis-radar.net/i2/1/13960/j1/20/74/1538721622/providers.json?imagesok=1&n=1&p=1&r=1&t=1
  (WARNING) 2018-10-04 23:41:39,037 inspector_log._OnMessage:26  DevTools console [network]: Failed to load resource: the server responded with a status of 404 (Not Found) https://i2-zzzzzzzzzzzzzzzzzbbbbbbbbbbbbb.init.cedexis-radar.net/i2/1/13960/j1/20/74/1538721622/providers.json?imagesok=1&n=1&p=1&r=1&t=1
  (WARNING) 2018-10-04 23:41:39,037 inspector_log._OnMessage:26  DevTools console [javascript]: Access to XMLHttpRequest at 'https://i2-zzzzzzzzzzzzzzzzzbbbbbbbbbbbbb.init.cedexis-radar.net/i2/1/13960/j1/20/74/1538721622/providers.json?imagesok=1&n=1&p=1&r=1&t=1' from origin 'https://www.tumblr.com' has been blocked by CORS policy: No 'Access-Control-Allow-Origin' header is present on the requested resource. https://www.tumblr.com/
  (ERROR) 2018-10-04 23:42:12,105 story_runner.ProcessError:100  Handleable error
  Traceback (most recent call last):
    File "/b/s/w/ir/third_party/catapult/telemetry/telemetry/internal/story_runner.py", line 126, in _RunStoryAndProcessErrorIfNeeded
      state.RunStory(results)
    File "/b/s/w/ir/third_party/catapult/common/py_trace_event/py_trace_event/trace_event_impl/decorators.py", line 75, in traced_function
      return func(*args, **kwargs)
    File "/b/s/w/ir/third_party/catapult/telemetry/telemetry/page/shared_page_state.py", line 300, in RunStory
      self._current_page.Run(self)
    File "/b/s/w/ir/third_party/catapult/telemetry/telemetry/page/__init__.py", line 95, in Run
      shared_state.page_test.RunNavigateSteps(self, current_tab)
    File "/b/s/w/ir/third_party/catapult/common/py_trace_event/py_trace_event/trace_event_impl/decorators.py", line 75, in traced_function
      return func(*args, **kwargs)
    File "/b/s/w/ir/third_party/catapult/telemetry/telemetry/page/legacy_page_test.py", line 152, in RunNavigateSteps
      page.RunNavigateSteps(action_runner)
    File "/b/s/w/ir/tools/perf/contrib/leak_detection/page_sets.py", line 29, in RunNavigateSteps
      self._WaitForPageLoadToComplete(new_tab.action_runner)
    File "/b/s/w/ir/tools/perf/contrib/leak_detection/page_sets.py", line 36, in _WaitForPageLoadToComplete
      py_utils.WaitFor(action_runner.tab.HasReachedQuiescence, timeout=30)
    File "/b/s/w/ir/third_party/catapult/common/py_utils/py_utils/__init__.py", line 148, in WaitFor
      (timeout, GetConditionString()))
  TimeoutException: Timed out while waiting 30s for HasReachedQuiescence.
  (WARNING) 2018-10-04 23:42:12,106 desktop_browser_backend._GetAllCrashpadMinidumps:322  No path to crashpad_database_util found
  (WARNING) 2018-10-04 23:42:12,107 shared_page_state.DumpStateUponFailure:130  Taking screenshots upon failures disabled.
  (ERROR) 2018-10-04 23:42:12,107 page_test_results.Fail:540  Failure recorded: Exception raised running https://www.tumblr.com/
  [  FAILED  ] memory.leak_detection/https://www.tumblr.com/ (40166 ms)

It's a timeout, so maybe just a flake?

This seems to be the first failure:
https://logs.chromium.org/logs/chromium/buildbucket/cr-buildbucket.appspot.com/8933568242309217264/+/steps/memory.leak_detection/0/stdout

Log excerpt:

  [ RUN      ] memory.leak_detection/http://www.livedoor.jp
  (ERROR) 2018-10-04 12:57:21,980 page_test_results.Fail:540  Failure recorded: Error: Memory leak is found.
      at new leakDetectionMetric (/tracing/metrics/blink/leak_detection_metric.html:50:15)
      at runMetrics (/b/s/w/ir/third_party/catapult/tracing/tracing/metrics/metric_map_function.html:61:16)
      at metricMapFunction (/b/s/w/ir/third_party/catapult/tracing/tracing/metrics/metric_map_function.html:190:24)
      at Object.mapSingleTrace (/tracing/mre/map_single_trace.html:39:7)
      at eval (/b/s/w/ir/third_party/catapult/tracing/tracing/mre/map_single_trace_cmdline.html:61:18)
      at Object.runAndConvertErrorsToFailures (/tracing/mre/map_single_trace.html:24:10)
      at mapSingleTraceWithResult (/b/s/w/ir/third_party/catapult/tracing/tracing/mre/map_single_trace_cmdline.html:52:12)
      at Object.mapSingleTraceMain (/b/s/w/ir/third_party/catapult/tracing/tracing/mre/map_single_trace_cmdline.html:77:20)
      at eval (/b/s/w/ir/third_party/catapult/tracing/tracing/mre/map_single_trace_cmdline.html:90:15)
      at Object.loadHTMLFile (/b/s/w/ir/third_party/catapult/third_party/vinn/vinn/html_imports_loader.js:140:5)
  [  FAILED  ] memory.leak_detection/http://www.livedoor.jp (11473 ms)


Still, can't relate to a commit from the V8 changes range.

There are 36 failures in a row on that bot now.

The excerpt in comment #3 (failure on livedoor.jp) only appeared in the first one (#21225). All the rest look like comment #2 (timeout on tumblr.com).

So the initial problem in #21225 has probably been fixed, and another one has been introduced in #21226.

https://ci.chromium.org/p/chromium/builders/luci.chromium.ci/Leak%20Detection%20Linux/21226

That one contains a SKIA roll, which could have caused the problem?

+sullivan@, you were overseeing this roll. Can you PTAL?

I tried running this locally and the tumblr.com page appears to be loaded, but the test just sits there until it times out.

I notice that the test has the following behaviour for some of the urls: https://cs.chromium.org/chromium/src/tools/perf/contrib/leak_detection/page_sets.py?rcl=09b186d5af7874816fe47bfcd91fbaab2476b0e6&l=39

When I use this for tumblr.com, the test passes. Perhaps it was a change to the page that is causing this?

Actually, I looked at the code for detecting NetworkQuiescence and I'm seeing nonsense values here:
https://cs.chromium.org/chromium/src/third_party/catapult/telemetry/telemetry/internal/browser/network_quiescence.js?rcl=de52d9ad295ecc6401de6d7cc3b468a0e8292308&l=97

lastResponseTimeMs is negative.

So the catapult roll seems like it could be the culprit.

CL to revert: https://chromium-review.googlesource.com/c/catapult/+/1265125

That revert is abandoned in favour of https://chromium-review.googlesource.com/c/chromium/src/+/1265128

Also I can confirm from reverting locally that the catapult CL is indeed the cause.

Note that when the test used to pass there was an extra warning which looks relevant to the catapult change:
(WARNING) 2018-10-05 15:51:51,330 inspector_log._OnMessage:26  DevTools console [security]: Refused to execute inline script because it violates the following Content Security Policy directive: "script-src 'self' https://assets.tumblr.com https://assets.tumblr.com https://sb.scorecardresearch.com https://*.google-analytics.com https://fc.yahoo.com https://s.yimg.com https://www.googletagservices.com/tag/js/gpt.js https://securepubads.g.doubleclick.net/gpt/ https://securepubads.g.doubleclick.net/gampad/ads https://pagead2.googlesyndication.com/pagead/osd.js https://www.google.com/recaptcha/api.js https://consent.cmp.oath.com https://*.cedexis.com https://*.cedexis-test.com https://*.gemini.yahoo.com 'unsafe-eval' 'nonce-BFHwgl37wZXUyZB3ZTjn4PZxu4' https://www.google.com/recaptcha/api.js https://www.gstatic.com". Either the 'unsafe-inline' keyword, a hash ('sha256-q7YJ/IoLUtQJvBbS70pIPtfe3uOAppNu6iy1w45CYzA='), or a nonce ('nonce-...') is required to enable inline execution.
 https://www.tumblr.com/

The following revision refers to this bug:
  https://chromium.googlesource.com/chromium/src.git/+/06487dd557d9f6da6f1b13de0babdb31e23b90a2

commit 06487dd557d9f6da6f1b13de0babdb31e23b90a2
Author: Kevin McNee <mcnee@chromium.org>
Date: Fri Oct 05 21:31:25 2018

Disable www.tumblr.com in memory.leak_detection

Loading this page is timing out due to a catapult roll.

TBR=yuzus@chromium.org

Bug: 892352
Change-Id: I0493da4014439537ede38679db2e6ba702d01247
Reviewed-on: https://chromium-review.googlesource.com/c/1265128
Reviewed-by: Kevin McNee <mcnee@chromium.org>
Commit-Queue: Kevin McNee <mcnee@chromium.org>
Cr-Commit-Position: refs/heads/master@{#597308}
[modify] https://crrev.com/06487dd557d9f6da6f1b13de0babdb31e23b90a2/tools/perf/contrib/leak_detection/page_sets.py

Removing Sheriff-Chromium as the failing test is disabled.